mnist-digit-recognition.py

# %%
# pandas version should be >= 0.24.1
!pip install --upgrade pandas

# %%
import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from keras.utils.np_utils import to_categorical
from keras import models, layers, optimizers
from keras.models import Sequential
from keras.layers import Dense, Activation, Conv2D, Flatten, MaxPool2D, Dropout
from keras.optimizers import Adam
from keras import backend as K

assert(pd.__version__ > '0.24.1')

# %%
train = pd.read_csv('../input/train.csv')
test = pd.read_csv('../input/test.csv')

# %%
y = train['label'].to_numpy().reshape((train.shape[0], 1))
y = to_categorical(y, num_classes=10)

del train['label']
X = train.to_numpy()

pred = test.to_numpy().reshape((-1, 28, 28, 1)) / 255.0

# %%
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.15)

X_train = X_train / 255.0
X_test = X_test / 255.0

X_train = X_train.reshape((-1, 28, 28, 1))
X_test = X_test.reshape((-1, 28, 28, 1))

# %%
model = Sequential()

model.add(Conv2D(32, kernel_size=(5, 5), input_shape=(X_train.shape[1:]), activation='tanh'))
model.add(Conv2D(32, kernel_size=(5, 5), activation='tanh'))
model.add(Dropout(0.2))

model.add(Conv2D(64, kernel_size=(3, 3), activation='tanh'))
model.add(Conv2D(64, kernel_size=(3, 3), activation='tanh'))
model.add(Dropout(0.2))

model.add(Flatten())
model.add(Dense(512, activation='tanh'))
model.add(Dense(128, activation='tanh'))
model.add(Dense(10, activation='softmax'))

adam = Adam(lr=0.001)
model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])
model.summary()

# %%
model.fit(X_train, y_train, epochs=50)

# %%
model.evaluate(X_test, y_test, verbose=0)

# %%
predictions = np.argmax(model.predict(pred), axis=-1)
predictions = predictions.reshape((predictions.shape[0], 1))
submission = pd.DataFrame({'ImageId':[i for i in range(1, predictions.shape[0] +1)], 'Label':predictions[:,0]})
submission.to_csv('submission.csv', index=False)

# %%