DNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "DNN.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "[View in Colaboratory](https://colab.research.google.com/github/santhavathi/Deep-Learning-MLBLR/blob/master/DNN.ipynb)"
      ]
    },
    {
      "metadata": {
        "id": "58WM5b1jXPy9",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Dropout, Activation, Flatten, Add\n",
        "from keras.layers import Convolution2D, MaxPooling2D, SeparableConv2D\n",
        "from keras.utils import np_utils\n",
        "from keras.optimizers import SGD\n",
        "\n",
        "from keras.datasets import mnist"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "kxLHZFyhXw4V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "53c0b41a-f597-49e6-9a4c-65bebb4f3644"
      },
      "cell_type": "code",
      "source": [
        "(X_train, y_train), (X_test, y_test) = mnist.load_data()"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://s3.amazonaws.com/img-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 3s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "8McH1J0iXy8V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "outputId": "ac53efd0-a917-4c41-eb69-08d7abf1adac"
      },
      "cell_type": "code",
      "source": [
        "print (X_train.shape)\n",
        "from matplotlib import pyplot as plt\n",
        "%matplotlib inline\n",
        "plt.imshow(X_train[0])"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(60000, 28, 28)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f8132405c50>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPoAAAD4CAYAAADFJPs2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f815676a978>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "dGA71d-VeHnU",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "X_train = X_train.reshape(X_train.shape[0], 28, 28,1)\n",
        "X_test = X_test.reshape(X_test.shape[0], 28, 28,1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fGRH6CHQsSsd",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "X_train = X_train.astype('float32')\n",
        "X_test = X_test.astype('float32')\n",
        "X_train /= 255\n",
        "X_test /= 255"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "69bqpmR9sejX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "3e448710-699e-41ce-d10d-21660a46dae7"
      },
      "cell_type": "code",
      "source": [
        "y_train[:10]"
      ],
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([5, 0, 4, 1, 9, 2, 1, 3, 1, 4], dtype=uint8)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 57
        }
      ]
    },
    {
      "metadata": {
        "id": "-IVZJWaJsfNM",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "# Convert 1-dimensional class arrays to 10-dimensional class matrices\n",
        "Y_train = np_utils.to_categorical(y_train, 10)\n",
        "Y_test = np_utils.to_categorical(y_test, 10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "2_oRqQ7Gsirv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        },
        "outputId": "73eb0ec4-d1ce-488e-c520-de11e1b91bb6"
      },
      "cell_type": "code",
      "source": [
        "from keras.layers import Activation\n",
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(Convolution2D(10, 26))\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "pNxXggn2xjoc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 306
        },
        "outputId": "1dfee971-ad76-408b-f04c-b880ec71ea89"
      },
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_80 (Conv2D)           (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_81 (Conv2D)           (None, 26, 26, 10)        330       \n",
            "_________________________________________________________________\n",
            "conv2d_82 (Conv2D)           (None, 1, 1, 10)          67610     \n",
            "_________________________________________________________________\n",
            "flatten_14 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_9 (Activation)    (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 68,260\n",
            "Trainable params: 68,260\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "KsKrftxBsm_I",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Q4RIyydGxvM1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "97fb959d-a967-4cf9-a7f0-31e8cbf310b9"
      },
      "cell_type": "code",
      "source": [
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 39s 655us/step - loss: 0.1902 - acc: 0.9455\n",
            "Epoch 2/10\n",
            "19168/60000 [========>.....................] - ETA: 26s - loss: 0.0654 - acc: 0.9806"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 649us/step - loss: 0.0671 - acc: 0.9794\n",
            "Epoch 3/10\n",
            "32672/60000 [===============>..............] - ETA: 17s - loss: 0.0494 - acc: 0.9848"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 651us/step - loss: 0.0498 - acc: 0.9844\n",
            "Epoch 4/10\n",
            "38592/60000 [==================>...........] - ETA: 13s - loss: 0.0359 - acc: 0.9892"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 651us/step - loss: 0.0390 - acc: 0.9876\n",
            "Epoch 5/10\n",
            "41344/60000 [===================>..........] - ETA: 12s - loss: 0.0280 - acc: 0.9911"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 647us/step - loss: 0.0298 - acc: 0.9905\n",
            "Epoch 6/10\n",
            "42464/60000 [====================>.........] - ETA: 11s - loss: 0.0229 - acc: 0.9923"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 646us/step - loss: 0.0232 - acc: 0.9923\n",
            "Epoch 7/10\n",
            "42816/60000 [====================>.........] - ETA: 11s - loss: 0.0194 - acc: 0.9932"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 648us/step - loss: 0.0203 - acc: 0.9930\n",
            "Epoch 8/10\n",
            "43072/60000 [====================>.........] - ETA: 11s - loss: 0.0136 - acc: 0.9956"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 654us/step - loss: 0.0148 - acc: 0.9953\n",
            "Epoch 9/10\n",
            "43232/60000 [====================>.........] - ETA: 10s - loss: 0.0122 - acc: 0.9960"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 652us/step - loss: 0.0129 - acc: 0.9959\n",
            "Epoch 10/10\n",
            "43232/60000 [====================>.........] - ETA: 10s - loss: 0.0086 - acc: 0.9971"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 646us/step - loss: 0.0098 - acc: 0.9969\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fd54f0b9c88>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 56
        }
      ]
    },
    {
      "metadata": {
        "id": "-vMC5RKOsxCm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "bd60d516-3239-4a05-b0f2-5702f0b08fff"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 57,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.10257896338440778, 0.978]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "p19HJXgXs7td",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "y_pred = model.predict(X_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "sebolUzhtAkp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "outputId": "1b68e70d-6e09-425f-abc9-0ecf747a2260"
      },
      "cell_type": "code",
      "source": [
        "print(y_pred[:9])\n",
        "print(y_test[:9])"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[1.00127746e-33 2.37473993e-29 6.33821458e-20 3.14192516e-09\n",
            "  7.90402717e-36 3.99499949e-24 0.00000000e+00 1.00000000e+00\n",
            "  3.57610068e-15 9.97443658e-16]\n",
            " [8.73376272e-19 1.05356175e-17 1.00000000e+00 3.33501885e-17\n",
            "  1.53687634e-31 6.15205749e-25 1.65097070e-12 1.13418687e-37\n",
            "  4.83411431e-17 5.77895863e-29]\n",
            " [6.00833619e-11 9.99997735e-01 1.05736231e-09 5.59661302e-12\n",
            "  2.08753158e-06 2.85256878e-14 7.84474916e-12 2.56375348e-08\n",
            "  6.43905835e-08 5.48774984e-11]\n",
            " [1.00000000e+00 2.17430465e-26 7.52944929e-11 4.01834446e-21\n",
            "  2.41257405e-25 8.41057850e-22 1.30504166e-16 1.31374079e-19\n",
            "  6.27372902e-18 6.73751764e-16]\n",
            " [5.71484060e-22 6.06575305e-24 5.42644881e-16 3.91771088e-13\n",
            "  9.99997377e-01 2.31268925e-22 1.39631117e-21 1.29384743e-13\n",
            "  4.69449271e-11 2.64376581e-06]\n",
            " [1.31096917e-16 9.99999881e-01 7.22091115e-13 1.32282165e-13\n",
            "  2.24967101e-10 7.99511714e-22 1.48020045e-15 1.45312242e-07\n",
            "  4.57901361e-09 2.59416273e-12]\n",
            " [2.12250745e-28 1.81931106e-20 1.87775034e-12 5.77876286e-19\n",
            "  9.99755085e-01 9.19374855e-16 2.75741469e-29 5.60949052e-15\n",
            "  2.42862763e-04 1.99777651e-06]\n",
            " [0.00000000e+00 2.46592043e-15 1.13774129e-20 7.18160127e-17\n",
            "  6.23333416e-13 8.99319783e-15 3.29248268e-32 1.37988796e-23\n",
            "  1.02933028e-14 1.00000000e+00]\n",
            " [1.47294712e-17 7.88286371e-31 3.71186597e-27 2.72067398e-22\n",
            "  6.57132918e-21 7.63899148e-01 2.36078218e-01 1.05309823e-31\n",
            "  2.26551019e-05 2.71181818e-12]]\n",
            "[7 2 1 0 4 1 4 9 5]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "_t3RIb48yoYK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 462
        },
        "outputId": "082de3a2-7c0b-4ff6-e5e1-43edfe452c4f"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(16, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 12))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 80,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_105 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_106 (Conv2D)          (None, 24, 24, 16)        4624      \n",
            "_________________________________________________________________\n",
            "conv2d_107 (Conv2D)          (None, 24, 24, 16)        272       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_19 (MaxPooling (None, 12, 12, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_108 (Conv2D)          (None, 1, 1, 10)          23050     \n",
            "_________________________________________________________________\n",
            "flatten_22 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_17 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 28,266\n",
            "Trainable params: 28,266\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "PHQ4Dm-G0Add",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "78feec17-ea51-4334-81f3-2dc71f806f7e"
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 81,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 86s 1ms/step - loss: 0.1674 - acc: 0.9496\n",
            "Epoch 2/10\n",
            "  736/60000 [..............................] - ETA: 1:23 - loss: 0.0607 - acc: 0.9810"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0648 - acc: 0.9800\n",
            "Epoch 3/10\n",
            " 5152/60000 [=>............................] - ETA: 1:16 - loss: 0.0495 - acc: 0.9839"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 84s 1ms/step - loss: 0.0486 - acc: 0.9850\n",
            "Epoch 4/10\n",
            " 7424/60000 [==>...........................] - ETA: 1:13 - loss: 0.0365 - acc: 0.9891"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0381 - acc: 0.9882\n",
            "Epoch 5/10\n",
            " 7936/60000 [==>...........................] - ETA: 1:13 - loss: 0.0267 - acc: 0.9918"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0309 - acc: 0.9906\n",
            "Epoch 6/10\n",
            " 8224/60000 [===>..........................] - ETA: 1:12 - loss: 0.0173 - acc: 0.9946"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 84s 1ms/step - loss: 0.0260 - acc: 0.9917\n",
            "Epoch 7/10\n",
            " 8480/60000 [===>..........................] - ETA: 1:11 - loss: 0.0167 - acc: 0.9942"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0217 - acc: 0.9930\n",
            "Epoch 8/10\n",
            " 8352/60000 [===>..........................] - ETA: 1:13 - loss: 0.0175 - acc: 0.9944"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 86s 1ms/step - loss: 0.0181 - acc: 0.9940\n",
            "Epoch 9/10\n",
            " 8224/60000 [===>..........................] - ETA: 1:14 - loss: 0.0148 - acc: 0.9955"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 86s 1ms/step - loss: 0.0165 - acc: 0.9943\n",
            "Epoch 10/10\n",
            " 7776/60000 [==>...........................] - ETA: 1:14 - loss: 0.0094 - acc: 0.9970"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 86s 1ms/step - loss: 0.0132 - acc: 0.9954\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f8134f0d518>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 81
        }
      ]
    },
    {
      "metadata": {
        "id": "o0ege40D0b6T",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "9eb2648e-57e2-4a5c-b267-1de910998742"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.05706005846480202, 0.9874]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "AFJmumhV9stj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 462
        },
        "outputId": "abeba538-4f8c-4bd1-8a58-5064fcc950c3"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(64, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(16, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 12))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 86,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_121 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_122 (Conv2D)          (None, 24, 24, 64)        18496     \n",
            "_________________________________________________________________\n",
            "conv2d_123 (Conv2D)          (None, 24, 24, 16)        1040      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_23 (MaxPooling (None, 12, 12, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_124 (Conv2D)          (None, 1, 1, 10)          23050     \n",
            "_________________________________________________________________\n",
            "flatten_26 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_21 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 42,906\n",
            "Trainable params: 42,906\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(64, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "cX-5p1OR_ixs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "2c11679f-94d9-4427-aecb-44b8e3abd1f0"
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.1484 - acc: 0.9546"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 150s 2ms/step - loss: 0.1483 - acc: 0.9546\n",
            "Epoch 2/10\n",
            "11808/60000 [====>.........................] - ETA: 2:00 - loss: 0.0527 - acc: 0.9846"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 150s 3ms/step - loss: 0.0544 - acc: 0.9835\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 3/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0408 - acc: 0.9874"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 150s 2ms/step - loss: 0.0408 - acc: 0.9874\n",
            "Epoch 4/10\n",
            "11808/60000 [====>.........................] - ETA: 2:00 - loss: 0.0374 - acc: 0.9887"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 150s 3ms/step - loss: 0.0328 - acc: 0.9893\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 5/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0263 - acc: 0.9920"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 151s 3ms/step - loss: 0.0263 - acc: 0.9920\n",
            "Epoch 6/10\n",
            "11808/60000 [====>.........................] - ETA: 2:00 - loss: 0.0191 - acc: 0.9936"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 150s 3ms/step - loss: 0.0222 - acc: 0.9928\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 7/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0181 - acc: 0.9940"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 151s 3ms/step - loss: 0.0182 - acc: 0.9940\n",
            "Epoch 8/10\n",
            "11776/60000 [====>.........................] - ETA: 2:00 - loss: 0.0137 - acc: 0.9962"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 150s 3ms/step - loss: 0.0164 - acc: 0.9948\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 9/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0138 - acc: 0.9953"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 150s 3ms/step - loss: 0.0138 - acc: 0.9953\n",
            "Epoch 10/10\n",
            "11744/60000 [====>.........................] - ETA: 2:01 - loss: 0.0092 - acc: 0.9968"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 151s 3ms/step - loss: 0.0116 - acc: 0.9959\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f8130438da0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 87
        }
      ]
    },
    {
      "metadata": {
        "id": "Np5aPrTzEd4l",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "2eb58203-7cee-455e-ff2a-77fcb78d58c3"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 88,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.0540941081030338, 0.9861]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "qgErFVjaDwXh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 394
        },
        "outputId": "9b86179f-054e-46b2-e8d6-5990ee75be83"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "#model.add(Convolution2D(32, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(16, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 13))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_129 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_130 (Conv2D)          (None, 26, 26, 16)        528       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_25 (MaxPooling (None, 13, 13, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_131 (Conv2D)          (None, 1, 1, 10)          27050     \n",
            "_________________________________________________________________\n",
            "flatten_28 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_23 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 27,898\n",
            "Trainable params: 27,898\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "G7uzbHJaF9nE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "39e429cf-4d81-41fe-e81d-5504352716e4"
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 40s 674us/step - loss: 0.2173 - acc: 0.9371\n",
            "Epoch 2/10\n",
            "18784/60000 [========>.....................] - ETA: 27s - loss: 0.0826 - acc: 0.9757"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 665us/step - loss: 0.0786 - acc: 0.9764\n",
            "Epoch 3/10\n",
            "31744/60000 [==============>...............] - ETA: 18s - loss: 0.0608 - acc: 0.9817"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 665us/step - loss: 0.0621 - acc: 0.9811\n",
            "Epoch 4/10\n",
            "37696/60000 [=================>............] - ETA: 14s - loss: 0.0515 - acc: 0.9832"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 662us/step - loss: 0.0515 - acc: 0.9836\n",
            "Epoch 5/10\n",
            "40416/60000 [===================>..........] - ETA: 12s - loss: 0.0433 - acc: 0.9866"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 662us/step - loss: 0.0438 - acc: 0.9863\n",
            "Epoch 6/10\n",
            "41952/60000 [===================>..........] - ETA: 11s - loss: 0.0353 - acc: 0.9886"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 638us/step - loss: 0.0359 - acc: 0.9886\n",
            "Epoch 7/10\n",
            "42272/60000 [====================>.........] - ETA: 11s - loss: 0.0310 - acc: 0.9896"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 662us/step - loss: 0.0319 - acc: 0.9896\n",
            "Epoch 8/10\n",
            "41824/60000 [===================>..........] - ETA: 12s - loss: 0.0257 - acc: 0.9917"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 674us/step - loss: 0.0264 - acc: 0.9914\n",
            "Epoch 9/10\n",
            "40608/60000 [===================>..........] - ETA: 13s - loss: 0.0223 - acc: 0.9927"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 674us/step - loss: 0.0238 - acc: 0.9924\n",
            "Epoch 10/10\n",
            "41728/60000 [===================>..........] - ETA: 11s - loss: 0.0196 - acc: 0.9936"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 629us/step - loss: 0.0207 - acc: 0.9932\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f81313b2dd8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 93
        }
      ]
    },
    {
      "metadata": {
        "id": "Wlhwd4spH3lC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "db95683c-62cc-4c98-a673-15fd58cdfd3d"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.06246836629922036, 0.9836]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "a0U6-Pt-IFtu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 462
        },
        "outputId": "fec7e736-cd6b-407c-9d2b-d3087a8c4461"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(3,3))\n",
        "model.add(Convolution2D(10, 8))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 99,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_143 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_144 (Conv2D)          (None, 24, 24, 16)        4624      \n",
            "_________________________________________________________________\n",
            "conv2d_145 (Conv2D)          (None, 24, 24, 10)        170       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_29 (MaxPooling (None, 8, 8, 10)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_146 (Conv2D)          (None, 1, 1, 10)          6410      \n",
            "_________________________________________________________________\n",
            "flatten_30 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_25 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 11,524\n",
            "Trainable params: 11,524\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "BvnDhjeKIOoQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "0f69edcf-3036-42c2-bda6-3e1a1bed662b"
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 100,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 87s 1ms/step - loss: 0.2389 - acc: 0.9247\n",
            "Epoch 2/10\n",
            "  832/60000 [..............................] - ETA: 1:25 - loss: 0.0864 - acc: 0.9712"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0733 - acc: 0.9777\n",
            "Epoch 3/10\n",
            " 5056/60000 [=>............................] - ETA: 1:18 - loss: 0.0553 - acc: 0.9842"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 86s 1ms/step - loss: 0.0553 - acc: 0.9830\n",
            "Epoch 4/10\n",
            " 6688/60000 [==>...........................] - ETA: 1:16 - loss: 0.0424 - acc: 0.9865"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 86s 1ms/step - loss: 0.0463 - acc: 0.9854\n",
            "Epoch 5/10\n",
            " 7552/60000 [==>...........................] - ETA: 1:14 - loss: 0.0380 - acc: 0.9886"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0401 - acc: 0.9875\n",
            "Epoch 6/10\n",
            " 7712/60000 [==>...........................] - ETA: 1:15 - loss: 0.0293 - acc: 0.9898"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0360 - acc: 0.9887\n",
            "Epoch 7/10\n",
            " 8448/60000 [===>..........................] - ETA: 1:13 - loss: 0.0281 - acc: 0.9911"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0332 - acc: 0.9895\n",
            "Epoch 8/10\n",
            " 8384/60000 [===>..........................] - ETA: 1:13 - loss: 0.0222 - acc: 0.9925"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0305 - acc: 0.9903\n",
            "Epoch 9/10\n",
            " 8480/60000 [===>..........................] - ETA: 1:12 - loss: 0.0324 - acc: 0.9903"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0286 - acc: 0.9908\n",
            "Epoch 10/10\n",
            " 8640/60000 [===>..........................] - ETA: 1:12 - loss: 0.0222 - acc: 0.9935"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0257 - acc: 0.9920\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f812f026668>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 100
        }
      ]
    },
    {
      "metadata": {
        "id": "TfKf2DrpN3ic",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "8eeda59a-70c8-4ee3-ace9-cec02918bc8b"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 102,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.04704412840785517, 0.9869]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "ah-vVke7Quno",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 530
        },
        "outputId": "9d3a179c-3593-49a2-bef9-fb351daa656a"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(8, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 11))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 103,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_147 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_148 (Conv2D)          (None, 24, 24, 16)        4624      \n",
            "_________________________________________________________________\n",
            "conv2d_149 (Conv2D)          (None, 22, 22, 8)         1160      \n",
            "_________________________________________________________________\n",
            "conv2d_150 (Conv2D)          (None, 22, 22, 10)        90        \n",
            "_________________________________________________________________\n",
            "max_pooling2d_30 (MaxPooling (None, 11, 11, 10)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_151 (Conv2D)          (None, 1, 1, 10)          12110     \n",
            "_________________________________________________________________\n",
            "flatten_31 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_26 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 18,304\n",
            "Trainable params: 18,304\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:6: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(8, (3, 3), activation=\"relu\")`\n",
            "  \n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "lFMNZJ85RWGh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "aef9f996-efc0-4c06-9969-f17a3eaada0f"
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.1928 - acc: 0.9428"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 110s 2ms/step - loss: 0.1927 - acc: 0.9428\n",
            "Epoch 2/10\n",
            "11808/60000 [====>.........................] - ETA: 1:28 - loss: 0.0808 - acc: 0.9761"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 111s 2ms/step - loss: 0.0763 - acc: 0.9771\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 3/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0582 - acc: 0.9822"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 110s 2ms/step - loss: 0.0582 - acc: 0.9822\n",
            "Epoch 4/10\n",
            "11808/60000 [====>.........................] - ETA: 1:28 - loss: 0.0497 - acc: 0.9840"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 111s 2ms/step - loss: 0.0488 - acc: 0.9851\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 5/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0421 - acc: 0.9868"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 111s 2ms/step - loss: 0.0421 - acc: 0.9868\n",
            "Epoch 6/10\n",
            "11808/60000 [====>.........................] - ETA: 1:28 - loss: 0.0315 - acc: 0.9900"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 111s 2ms/step - loss: 0.0366 - acc: 0.9887\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 7/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0319 - acc: 0.9901"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 111s 2ms/step - loss: 0.0319 - acc: 0.9902\n",
            "Epoch 8/10\n",
            "11808/60000 [====>.........................] - ETA: 1:28 - loss: 0.0265 - acc: 0.9920"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 111s 2ms/step - loss: 0.0288 - acc: 0.9908\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 9/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0246 - acc: 0.9917"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 111s 2ms/step - loss: 0.0246 - acc: 0.9917\n",
            "Epoch 10/10\n",
            "11744/60000 [====>.........................] - ETA: 1:30 - loss: 0.0156 - acc: 0.9953"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 112s 2ms/step - loss: 0.0231 - acc: 0.9925\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f812e40eeb8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 104
        }
      ]
    },
    {
      "metadata": {
        "id": "MAOnFsxZVb7S",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "064423aa-23e5-4194-87f1-062a4a59105e"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 105,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.05274732987669122, 0.9861]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "VBmdZGScV9a8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 462
        },
        "outputId": "c84962e8-5c0b-4e21-b537-decc8a14cd0c"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(32, 3, 3, activation='relu'))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 12))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 106,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_152 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_153 (Conv2D)          (None, 24, 24, 32)        9248      \n",
            "_________________________________________________________________\n",
            "conv2d_154 (Conv2D)          (None, 24, 24, 10)        330       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_31 (MaxPooling (None, 12, 12, 10)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_155 (Conv2D)          (None, 1, 1, 10)          14410     \n",
            "_________________________________________________________________\n",
            "flatten_32 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_27 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 24,308\n",
            "Trainable params: 24,308\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "TXJM9uCTWJXw",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "3e6b8dd6-6fd0-4489-9ad3-dd5ddc6599c0"
      },
      "cell_type": "code",
      "source": [
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer='adam',\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 107,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.2054 - acc: 0.9389"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 114s 2ms/step - loss: 0.2053 - acc: 0.9389\n",
            "Epoch 2/10\n",
            "11808/60000 [====>.........................] - ETA: 1:30 - loss: 0.0728 - acc: 0.9779"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 113s 2ms/step - loss: 0.0665 - acc: 0.9803\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 3/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0511 - acc: 0.9839"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 112s 2ms/step - loss: 0.0511 - acc: 0.9839\n",
            "Epoch 4/10\n",
            "11808/60000 [====>.........................] - ETA: 1:28 - loss: 0.0378 - acc: 0.9878"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 112s 2ms/step - loss: 0.0406 - acc: 0.9871\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 5/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0344 - acc: 0.9890"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 112s 2ms/step - loss: 0.0344 - acc: 0.9890\n",
            "Epoch 6/10\n",
            "11808/60000 [====>.........................] - ETA: 1:29 - loss: 0.0242 - acc: 0.9921"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 112s 2ms/step - loss: 0.0293 - acc: 0.9904\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 7/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0257 - acc: 0.9917"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 112s 2ms/step - loss: 0.0257 - acc: 0.9917\n",
            "Epoch 8/10\n",
            "11808/60000 [====>.........................] - ETA: 1:30 - loss: 0.0163 - acc: 0.9947"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 112s 2ms/step - loss: 0.0210 - acc: 0.9933\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 9/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0192 - acc: 0.9935"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 112s 2ms/step - loss: 0.0192 - acc: 0.9935\n",
            "Epoch 10/10\n",
            "11744/60000 [====>.........................] - ETA: 1:30 - loss: 0.0165 - acc: 0.9957"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 112s 2ms/step - loss: 0.0173 - acc: 0.9946\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f812e9b44e0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 107
        }
      ]
    },
    {
      "metadata": {
        "id": "cBBfgk5FYW9j",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "25a96c12-8cc1-472d-fb95-1b6a67ac4ef4"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.055117777645571915, 0.9855]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "AoOEChZBYXhQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 496
        },
        "outputId": "34acbb79-2e4f-4ce0-f475-a11f726f9010"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu'))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(16, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 12))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 111,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_162 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_163 (Conv2D)          (None, 24, 24, 16)        4624      \n",
            "_________________________________________________________________\n",
            "dropout_11 (Dropout)         (None, 24, 24, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_164 (Conv2D)          (None, 24, 24, 16)        272       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_34 (MaxPooling (None, 12, 12, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_165 (Conv2D)          (None, 1, 1, 10)          23050     \n",
            "_________________________________________________________________\n",
            "flatten_35 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_30 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 28,266\n",
            "Trainable params: 28,266\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "aQugJR5OYt09",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "c746023a-ef1c-4ca2-cffc-a27f5a59afa4"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 112,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "45728/60000 [=====================>........] - ETA: 22s - loss: 0.2834 - acc: 0.9155"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 94s 2ms/step - loss: 0.2420 - acc: 0.9278\n",
            "Epoch 2/10\n",
            "17440/60000 [=======>......................] - ETA: 1:05 - loss: 0.0912 - acc: 0.9732"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 93s 2ms/step - loss: 0.0821 - acc: 0.9748\n",
            "Epoch 3/10\n",
            " 5024/60000 [=>............................] - ETA: 1:25 - loss: 0.0550 - acc: 0.9841"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 93s 2ms/step - loss: 0.0622 - acc: 0.9804\n",
            "Epoch 4/10"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "49568/60000 [=======================>......] - ETA: 16s - loss: 0.0521 - acc: 0.9844"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 96s 2ms/step - loss: 0.0517 - acc: 0.9845\n",
            "Epoch 5/10\n",
            "13568/60000 [=====>........................] - ETA: 1:15 - loss: 0.0484 - acc: 0.9853"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 97s 2ms/step - loss: 0.0468 - acc: 0.9858\n",
            "Epoch 6/10\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "45760/60000 [=====================>........] - ETA: 23s - loss: 0.0424 - acc: 0.9867"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 97s 2ms/step - loss: 0.0420 - acc: 0.9870\n",
            "Epoch 7/10\n",
            "12768/60000 [=====>........................] - ETA: 1:15 - loss: 0.0377 - acc: 0.9886"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 97s 2ms/step - loss: 0.0384 - acc: 0.9879\n",
            "Epoch 8/10\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "43872/60000 [====================>.........] - ETA: 26s - loss: 0.0336 - acc: 0.9889"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 97s 2ms/step - loss: 0.0354 - acc: 0.9887\n",
            "Epoch 9/10\n",
            "11264/60000 [====>.........................] - ETA: 1:19 - loss: 0.0293 - acc: 0.9903"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "58752/60000 [============================>.] - ETA: 2s - loss: 0.0330 - acc: 0.9894"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 97s 2ms/step - loss: 0.0332 - acc: 0.9894\n",
            "Epoch 10/10\n",
            "17504/60000 [=======>......................] - ETA: 1:08 - loss: 0.0300 - acc: 0.9900"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 97s 2ms/step - loss: 0.0316 - acc: 0.9897\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f8130cf1ef0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 112
        }
      ]
    },
    {
      "metadata": {
        "id": "-jzWk64mgRCt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "daa2855a-1fae-4fb1-ee1e-2023bb12043d"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 113,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.03550854518567212, 0.9881]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "HZGs1iXIhPt4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "1b5c46dd-504e-4601-a8b1-f778f95c6ae0"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 1, activation='relu'))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 13))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 114,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_166 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_167 (Conv2D)          (None, 26, 26, 16)        528       \n",
            "_________________________________________________________________\n",
            "dropout_12 (Dropout)         (None, 26, 26, 16)        0         \n",
            "_________________________________________________________________\n",
            "max_pooling2d_35 (MaxPooling (None, 13, 13, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_168 (Conv2D)          (None, 1, 1, 10)          27050     \n",
            "_________________________________________________________________\n",
            "flatten_36 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_31 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 27,898\n",
            "Trainable params: 27,898\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "UU6NjcshhZd5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "3eef9523-5a85-442d-eb95-5f34d9435e7a"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 115,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 49s 819us/step - loss: 0.2424 - acc: 0.9263\n",
            "Epoch 2/10\n",
            " 6272/60000 [==>...........................] - ETA: 43s - loss: 0.1034 - acc: 0.9707"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 49s 810us/step - loss: 0.0917 - acc: 0.9721\n",
            "Epoch 3/10\n",
            "11872/60000 [====>.........................] - ETA: 39s - loss: 0.0718 - acc: 0.9771"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 48s 806us/step - loss: 0.0730 - acc: 0.9773\n",
            "Epoch 4/10\n",
            "15296/60000 [======>.......................] - ETA: 35s - loss: 0.0636 - acc: 0.9805"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 48s 807us/step - loss: 0.0632 - acc: 0.9808\n",
            "Epoch 5/10\n",
            "16064/60000 [=======>......................] - ETA: 35s - loss: 0.0519 - acc: 0.9835"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 49s 811us/step - loss: 0.0568 - acc: 0.9824\n",
            "Epoch 6/10\n",
            "16448/60000 [=======>......................] - ETA: 35s - loss: 0.0479 - acc: 0.9853"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 48s 808us/step - loss: 0.0501 - acc: 0.9846\n",
            "Epoch 7/10\n",
            "16832/60000 [=======>......................] - ETA: 34s - loss: 0.0473 - acc: 0.9855"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 49s 810us/step - loss: 0.0453 - acc: 0.9857\n",
            "Epoch 8/10\n",
            "16512/60000 [=======>......................] - ETA: 35s - loss: 0.0366 - acc: 0.9881"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 48s 807us/step - loss: 0.0431 - acc: 0.9861\n",
            "Epoch 9/10\n",
            "16832/60000 [=======>......................] - ETA: 35s - loss: 0.0413 - acc: 0.9868"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 49s 813us/step - loss: 0.0396 - acc: 0.9874\n",
            "Epoch 10/10\n",
            "16192/60000 [=======>......................] - ETA: 35s - loss: 0.0370 - acc: 0.9876"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 48s 800us/step - loss: 0.0374 - acc: 0.9881\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f812dfe9f98>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 115
        }
      ]
    },
    {
      "metadata": {
        "id": "KSn3uLoyjW-a",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "624495e6-9a06-4a31-823e-122dc119b5fd"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 116,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.04866545837819576, 0.9845]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "-qIT24NTlTHm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 530
        },
        "outputId": "b33869f4-51c0-47a7-be90-a9f6d2249e24"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu'))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(16, 1, activation='relu'))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 12))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 117,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_169 (Conv2D)          (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "conv2d_170 (Conv2D)          (None, 24, 24, 16)        4624      \n",
            "_________________________________________________________________\n",
            "dropout_13 (Dropout)         (None, 24, 24, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_171 (Conv2D)          (None, 24, 24, 16)        272       \n",
            "_________________________________________________________________\n",
            "dropout_14 (Dropout)         (None, 24, 24, 16)        0         \n",
            "_________________________________________________________________\n",
            "max_pooling2d_36 (MaxPooling (None, 12, 12, 16)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_172 (Conv2D)          (None, 1, 1, 10)          23050     \n",
            "_________________________________________________________________\n",
            "flatten_37 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_32 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 28,266\n",
            "Trainable params: 28,266\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "xl93agvdlnvs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "dfff79fd-9a55-4e81-f681-273cd3685f13"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 118,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.2256 - acc: 0.9326"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 103s 2ms/step - loss: 0.2255 - acc: 0.9327\n",
            "Epoch 2/10\n",
            "11808/60000 [====>.........................] - ETA: 1:22 - loss: 0.0905 - acc: 0.9717"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 102s 2ms/step - loss: 0.0852 - acc: 0.9735\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 3/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0664 - acc: 0.9794"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 102s 2ms/step - loss: 0.0663 - acc: 0.9794\n",
            "Epoch 4/10\n",
            "11808/60000 [====>.........................] - ETA: 1:22 - loss: 0.0604 - acc: 0.9813"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 102s 2ms/step - loss: 0.0582 - acc: 0.9815\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 5/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0513 - acc: 0.9839"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 101s 2ms/step - loss: 0.0513 - acc: 0.9839\n",
            "Epoch 6/10\n",
            "11808/60000 [====>.........................] - ETA: 1:21 - loss: 0.0479 - acc: 0.9837"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 103s 2ms/step - loss: 0.0494 - acc: 0.9841\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 7/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0451 - acc: 0.9858"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 103s 2ms/step - loss: 0.0451 - acc: 0.9858\n",
            "Epoch 8/10\n",
            "11808/60000 [====>.........................] - ETA: 1:22 - loss: 0.0354 - acc: 0.9881"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 103s 2ms/step - loss: 0.0411 - acc: 0.9867\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 9/10\n",
            "59968/60000 [============================>.] - ETA: 0s - loss: 0.0403 - acc: 0.9869"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\r60000/60000 [==============================] - 102s 2ms/step - loss: 0.0403 - acc: 0.9869\n",
            "Epoch 10/10\n",
            "11744/60000 [====>.........................] - ETA: 1:23 - loss: 0.0278 - acc: 0.9904"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 103s 2ms/step - loss: 0.0376 - acc: 0.9877\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f812cdd1ef0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 118
        }
      ]
    },
    {
      "metadata": {
        "id": "bChW6vDTljeS",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "176bd2f9-7107-4d04-a21e-89982087fc53"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 119,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.044542449210956694, 0.9867]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "kV07WLgcqLiE",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 462
        },
        "outputId": "ed9ff890-80e9-434e-91a2-218105d1252a"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(16, 3, 3, activation='relu', dilation_rate=2, input_shape=(28,28,1)))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu', dilation_rate=2))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 10))\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_17 (Conv2D)           (None, 24, 24, 16)        160       \n",
            "_________________________________________________________________\n",
            "conv2d_18 (Conv2D)           (None, 20, 20, 16)        2320      \n",
            "_________________________________________________________________\n",
            "conv2d_19 (Conv2D)           (None, 20, 20, 10)        170       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 10, 10, 10)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_20 (Conv2D)           (None, 1, 1, 10)          10010     \n",
            "_________________________________________________________________\n",
            "flatten_6 (Flatten)          (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_6 (Activation)    (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 12,660\n",
            "Trainable params: 12,660\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\", dilation_rate=2, input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\", dilation_rate=2)`\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "pWGw02QKuFa8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "6e2eb6a0-e58d-4d4e-97dd-533aaca490fd"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 52s 870us/step - loss: 0.2229 - acc: 0.9319\n",
            "Epoch 2/10\n",
            " 3744/60000 [>.............................] - ETA: 48s - loss: 0.1028 - acc: 0.9701"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 52s 861us/step - loss: 0.0794 - acc: 0.9761\n",
            "Epoch 3/10\n",
            " 8928/60000 [===>..........................] - ETA: 44s - loss: 0.0604 - acc: 0.9813"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 52s 864us/step - loss: 0.0585 - acc: 0.9820\n",
            "Epoch 4/10\n",
            "11296/60000 [====>.........................] - ETA: 41s - loss: 0.0496 - acc: 0.9844"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 52s 859us/step - loss: 0.0485 - acc: 0.9847\n",
            "Epoch 5/10\n",
            "12320/60000 [=====>........................] - ETA: 41s - loss: 0.0417 - acc: 0.9866"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 52s 867us/step - loss: 0.0419 - acc: 0.9867\n",
            "Epoch 6/10\n",
            "12704/60000 [=====>........................] - ETA: 40s - loss: 0.0310 - acc: 0.9896"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 51s 858us/step - loss: 0.0370 - acc: 0.9885\n",
            "Epoch 7/10\n",
            "12992/60000 [=====>........................] - ETA: 39s - loss: 0.0258 - acc: 0.9911"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 51s 845us/step - loss: 0.0316 - acc: 0.9901\n",
            "Epoch 8/10\n",
            "13152/60000 [=====>........................] - ETA: 40s - loss: 0.0279 - acc: 0.9916"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 51s 848us/step - loss: 0.0294 - acc: 0.9907\n",
            "Epoch 9/10\n",
            "13152/60000 [=====>........................] - ETA: 39s - loss: 0.0249 - acc: 0.9916"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 51s 851us/step - loss: 0.0262 - acc: 0.9911\n",
            "Epoch 10/10\n",
            "13120/60000 [=====>........................] - ETA: 39s - loss: 0.0216 - acc: 0.9934"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 51s 852us/step - loss: 0.0258 - acc: 0.9919\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fd5554c75f8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "metadata": {
        "id": "719X8kE9sVwP",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "14dd63e1-92d7-4238-b086-84c0af342043"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.042718834021483784, 0.9867]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "wL5u_Ig7PwOR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 530
        },
        "outputId": "fcf42fce-2e71-4a52-b138-1a2a8aec7407"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', dilation_rate=2, input_shape=(28,28,1)))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(10, 5))\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_43 (Conv2D)           (None, 24, 24, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_14 (MaxPooling (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_44 (Conv2D)           (None, 10, 10, 16)        4624      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_15 (MaxPooling (None, 5, 5, 16)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_45 (Conv2D)           (None, 5, 5, 10)          170       \n",
            "_________________________________________________________________\n",
            "dropout_6 (Dropout)          (None, 5, 5, 10)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_46 (Conv2D)           (None, 1, 1, 10)          2510      \n",
            "_________________________________________________________________\n",
            "flatten_12 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_12 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 7,624\n",
            "Trainable params: 7,624\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", dilation_rate=2, input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:6: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "rIevsOAQQRNg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "c6e8dfdd-e056-4675-f77a-edfb980bf863"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 38s 639us/step - loss: 0.3144 - acc: 0.8974\n",
            "Epoch 2/10\n",
            "19584/60000 [========>.....................] - ETA: 26s - loss: 0.1327 - acc: 0.9594"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 645us/step - loss: 0.1199 - acc: 0.9626\n",
            "Epoch 3/10\n",
            "32832/60000 [===============>..............] - ETA: 17s - loss: 0.0987 - acc: 0.9688"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 637us/step - loss: 0.0953 - acc: 0.9704\n",
            "Epoch 4/10\n",
            "39008/60000 [==================>...........] - ETA: 13s - loss: 0.0805 - acc: 0.9746"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 638us/step - loss: 0.0818 - acc: 0.9741\n",
            "Epoch 5/10\n",
            "41536/60000 [===================>..........] - ETA: 11s - loss: 0.0744 - acc: 0.9767"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 643us/step - loss: 0.0735 - acc: 0.9772\n",
            "Epoch 6/10\n",
            "42720/60000 [====================>.........] - ETA: 11s - loss: 0.0665 - acc: 0.9793"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 642us/step - loss: 0.0667 - acc: 0.9793\n",
            "Epoch 7/10\n",
            "43296/60000 [====================>.........] - ETA: 10s - loss: 0.0649 - acc: 0.9797"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 643us/step - loss: 0.0640 - acc: 0.9800\n",
            "Epoch 8/10\n",
            "43616/60000 [====================>.........] - ETA: 10s - loss: 0.0616 - acc: 0.9804"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 640us/step - loss: 0.0609 - acc: 0.9808\n",
            "Epoch 9/10\n",
            "43744/60000 [====================>.........] - ETA: 10s - loss: 0.0588 - acc: 0.9813"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 638us/step - loss: 0.0581 - acc: 0.9817\n",
            "Epoch 10/10\n",
            "43456/60000 [====================>.........] - ETA: 10s - loss: 0.0538 - acc: 0.9830"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 38s 639us/step - loss: 0.0540 - acc: 0.9830\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fd5537a2400>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "metadata": {
        "id": "m9xg-Ha-XoPs",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "b07dad62-cc83-4c8a-a7c8-d7104f288927"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.037724575306486804, 0.9882]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "exiOr2c9bXcn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 530
        },
        "outputId": "e6ede200-1a36-4f2a-8562-8e513b7d4672"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', dilation_rate=2, input_shape=(28,28,1)))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(SeparableConv2D(16, 3, 3, activation='relu'))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(10, 5))\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_51 (Conv2D)           (None, 24, 24, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_18 (MaxPooling (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "separable_conv2d_1 (Separabl (None, 10, 10, 16)        816       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_19 (MaxPooling (None, 5, 5, 16)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_52 (Conv2D)           (None, 5, 5, 10)          170       \n",
            "_________________________________________________________________\n",
            "dropout_8 (Dropout)          (None, 5, 5, 10)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_53 (Conv2D)           (None, 1, 1, 10)          2510      \n",
            "_________________________________________________________________\n",
            "flatten_14 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_14 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 3,816\n",
            "Trainable params: 3,816\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", dilation_rate=2, input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:6: UserWarning: Update your `SeparableConv2D` call to the Keras 2 API: `SeparableConv2D(16, (3, 3), activation=\"relu\")`\n",
            "  \n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "-4zd8XYHcfRX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "6bf89279-2dde-4e55-8a9d-e3010eb733ac"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 37s 615us/step - loss: 0.4361 - acc: 0.8557\n",
            "Epoch 2/10\n",
            "20096/60000 [=========>....................] - ETA: 24s - loss: 0.1433 - acc: 0.9562"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 611us/step - loss: 0.1388 - acc: 0.9575\n",
            "Epoch 3/10\n",
            "33312/60000 [===============>..............] - ETA: 16s - loss: 0.1198 - acc: 0.9618"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 612us/step - loss: 0.1177 - acc: 0.9627\n",
            "Epoch 4/10\n",
            "39360/60000 [==================>...........] - ETA: 12s - loss: 0.1083 - acc: 0.9672"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 611us/step - loss: 0.1075 - acc: 0.9674\n",
            "Epoch 5/10\n",
            "42144/60000 [====================>.........] - ETA: 10s - loss: 0.0976 - acc: 0.9698"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 611us/step - loss: 0.0980 - acc: 0.9695\n",
            "Epoch 6/10\n",
            "43328/60000 [====================>.........] - ETA: 10s - loss: 0.0913 - acc: 0.9715"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 613us/step - loss: 0.0901 - acc: 0.9718\n",
            "Epoch 7/10\n",
            "44000/60000 [=====================>........] - ETA: 9s - loss: 0.0855 - acc: 0.9723"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 612us/step - loss: 0.0845 - acc: 0.9728\n",
            "Epoch 8/10\n",
            "44224/60000 [=====================>........] - ETA: 9s - loss: 0.0808 - acc: 0.9742"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 609us/step - loss: 0.0808 - acc: 0.9746\n",
            "Epoch 9/10\n",
            "44160/60000 [=====================>........] - ETA: 9s - loss: 0.0802 - acc: 0.9744"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 616us/step - loss: 0.0795 - acc: 0.9749\n",
            "Epoch 10/10\n",
            "44032/60000 [=====================>........] - ETA: 9s - loss: 0.0752 - acc: 0.9767"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 37s 615us/step - loss: 0.0767 - acc: 0.9764\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fd5517b2ef0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "metadata": {
        "id": "-Qpa1Jqwd9tc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "3f2c4e45-b5d4-4164-b697-67c808d00fe9"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.04231452104093041, 0.9862]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "0ACclRRrrmgD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "outputId": "9d8bb952-e53b-4199-ede7-2b0fa53a86d0"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', dilation_rate=2, input_shape=(28,28,1)))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu', dilation_rate=2))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(Convolution2D(10, 4))\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", dilation_rate=2, input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:7: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\", dilation_rate=2)`\n",
            "  import sys\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_85 (Conv2D)           (None, 24, 24, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_38 (MaxPooling (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "dropout_18 (Dropout)         (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_86 (Conv2D)           (None, 8, 8, 16)          4624      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_39 (MaxPooling (None, 4, 4, 16)          0         \n",
            "_________________________________________________________________\n",
            "dropout_19 (Dropout)         (None, 4, 4, 16)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_87 (Conv2D)           (None, 4, 4, 10)          170       \n",
            "_________________________________________________________________\n",
            "conv2d_88 (Conv2D)           (None, 1, 1, 10)          1610      \n",
            "_________________________________________________________________\n",
            "flatten_25 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_25 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 6,724\n",
            "Trainable params: 6,724\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "rppbv1PFsHYc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "14cd2423-5119-44ac-b93e-11da521786fb"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 60,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 41s 677us/step - loss: 0.3677 - acc: 0.8827\n",
            "Epoch 2/10\n",
            "18368/60000 [========>.....................] - ETA: 27s - loss: 0.1432 - acc: 0.9545"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 659us/step - loss: 0.1332 - acc: 0.9590\n",
            "Epoch 3/10\n",
            "32064/60000 [===============>..............] - ETA: 18s - loss: 0.1097 - acc: 0.9659"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 658us/step - loss: 0.1063 - acc: 0.9673\n",
            "Epoch 4/10\n",
            "38048/60000 [==================>...........] - ETA: 14s - loss: 0.0942 - acc: 0.9706"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 663us/step - loss: 0.0942 - acc: 0.9705\n",
            "Epoch 5/10\n",
            "40448/60000 [===================>..........] - ETA: 12s - loss: 0.0864 - acc: 0.9745"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 659us/step - loss: 0.0867 - acc: 0.9735\n",
            "Epoch 6/10\n",
            "42048/60000 [====================>.........] - ETA: 11s - loss: 0.0791 - acc: 0.9745"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 654us/step - loss: 0.0807 - acc: 0.9742\n",
            "Epoch 7/10\n",
            "42528/60000 [====================>.........] - ETA: 11s - loss: 0.0775 - acc: 0.9758"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 655us/step - loss: 0.0757 - acc: 0.9766\n",
            "Epoch 8/10\n",
            "43104/60000 [====================>.........] - ETA: 11s - loss: 0.0716 - acc: 0.9775"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 653us/step - loss: 0.0742 - acc: 0.9770\n",
            "Epoch 9/10\n",
            "43488/60000 [====================>.........] - ETA: 10s - loss: 0.0683 - acc: 0.9793"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 649us/step - loss: 0.0684 - acc: 0.9790\n",
            "Epoch 10/10\n",
            "40128/60000 [===================>..........] - ETA: 14s - loss: 0.0719 - acc: 0.9775"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 42s 693us/step - loss: 0.0706 - acc: 0.9780\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fd54eda9240>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        }
      ]
    },
    {
      "metadata": {
        "id": "UP4ZNIlEtqNN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "845a89df-158f-4374-f697-9947c6220e89"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.0344324719496537, 0.9884]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "j3xV5Th4uTjc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 598
        },
        "outputId": "db608552-d8be-4b51-ef31-52edf4f26960"
      },
      "cell_type": "code",
      "source": [
        "model = Sequential()\n",
        "\n",
        " \n",
        "model.add(Convolution2D(32, 3, 3, activation='relu', dilation_rate=2, input_shape=(28,28,1)))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(16, 3, 3, activation='relu', dilation_rate=2))\n",
        "model.add(MaxPooling2D(2,2))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Convolution2D(10, 1, activation='relu'))\n",
        "model.add(Convolution2D(10, 4))\n",
        "model.add(Dropout(0.25))\n",
        "model.add(Flatten())\n",
        "model.add(Activation('softmax'))\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_97 (Conv2D)           (None, 24, 24, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_44 (MaxPooling (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "dropout_26 (Dropout)         (None, 12, 12, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_98 (Conv2D)           (None, 8, 8, 16)          4624      \n",
            "_________________________________________________________________\n",
            "max_pooling2d_45 (MaxPooling (None, 4, 4, 16)          0         \n",
            "_________________________________________________________________\n",
            "dropout_27 (Dropout)         (None, 4, 4, 16)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_99 (Conv2D)           (None, 4, 4, 10)          170       \n",
            "_________________________________________________________________\n",
            "conv2d_100 (Conv2D)          (None, 1, 1, 10)          1610      \n",
            "_________________________________________________________________\n",
            "dropout_28 (Dropout)         (None, 1, 1, 10)          0         \n",
            "_________________________________________________________________\n",
            "flatten_28 (Flatten)         (None, 10)                0         \n",
            "_________________________________________________________________\n",
            "activation_28 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 6,724\n",
            "Trainable params: 6,724\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(32, (3, 3), activation=\"relu\", dilation_rate=2, input_shape=(28, 28, 1...)`\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:7: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(16, (3, 3), activation=\"relu\", dilation_rate=2)`\n",
            "  import sys\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "metadata": {
        "id": "sFT4V_g-uiId",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 428
        },
        "outputId": "58be51b9-eecc-4e5e-c8fd-d60496bb7dd3"
      },
      "cell_type": "code",
      "source": [
        "sgd = SGD(lr=0.01, momentum=0.9)\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "             optimizer=sgd,\n",
        "             metrics=['accuracy'])\n",
        "model.fit(X_train, Y_train, batch_size=32, nb_epoch=10, verbose=1)"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/models.py:981: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n",
            "  warnings.warn('The `nb_epoch` argument in `fit` '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "60000/60000 [==============================] - 40s 675us/step - loss: 0.7365 - acc: 0.7215\n",
            "Epoch 2/10\n",
            "17952/60000 [=======>......................] - ETA: 28s - loss: 0.4856 - acc: 0.7973"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 43s 710us/step - loss: 0.4646 - acc: 0.8009\n",
            "Epoch 3/10\n",
            "23296/60000 [==========>...................] - ETA: 30s - loss: 0.4331 - acc: 0.8124"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 44s 733us/step - loss: 0.4263 - acc: 0.8115\n",
            "Epoch 4/10\n",
            "33888/60000 [===============>..............] - ETA: 17s - loss: 0.4174 - acc: 0.8129"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 659us/step - loss: 0.4159 - acc: 0.8130\n",
            "Epoch 5/10\n",
            "38400/60000 [==================>...........] - ETA: 14s - loss: 0.3956 - acc: 0.8174"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 663us/step - loss: 0.3966 - acc: 0.8164\n",
            "Epoch 6/10\n",
            "39840/60000 [==================>...........] - ETA: 13s - loss: 0.3913 - acc: 0.8194"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 666us/step - loss: 0.3933 - acc: 0.8168\n",
            "Epoch 7/10\n",
            "41440/60000 [===================>..........] - ETA: 12s - loss: 0.3838 - acc: 0.8188"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 40s 660us/step - loss: 0.3812 - acc: 0.8204\n",
            "Epoch 8/10\n",
            "42272/60000 [====================>.........] - ETA: 11s - loss: 0.3821 - acc: 0.8198"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 658us/step - loss: 0.3815 - acc: 0.8199\n",
            "Epoch 9/10\n",
            "42592/60000 [====================>.........] - ETA: 11s - loss: 0.3784 - acc: 0.8232"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 656us/step - loss: 0.3779 - acc: 0.8227\n",
            "Epoch 10/10\n",
            "42784/60000 [====================>.........] - ETA: 11s - loss: 0.3767 - acc: 0.8230"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 39s 650us/step - loss: 0.3746 - acc: 0.8234\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fd54eae49b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 66
        }
      ]
    },
    {
      "metadata": {
        "id": "Zgg0u4gEyzRT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "c617aff4-6d4d-43ff-8169-d18ffc56b937"
      },
      "cell_type": "code",
      "source": [
        "score = model.evaluate(X_test, Y_test, verbose=0)\n",
        "print(score)"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[0.038829236471117475, 0.9871]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "toYWMgKxeh0x",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "**Note:** Using 5x5 filter gives accuracy > 99%"
      ]
    }
  ]
}