Fish Species Classification Project

Project Purpose: The aim of this project is to develop a deep learning model that can accurately classify fish species. The model takes fish images as input and determines which species these images belong to.

Dataset: In the project, the "A Large Scale Fish Dataset" dataset on the Kaggle platform was used. This dataset contains images of 9 different fish species.

These fish species: •Hourse Mackerel •Black Sea Sprat •Sea Bass •Red Mullet •Trout •Striped Red Mullet •Shrimp •Gilt-Head Bream •Red Sea Bream

Used libraries: •Pandas: Used for data processing and manipulation. •Matplotlib & Seaborn: Used for visualization and data analysis. •NumPy: Used for matrix and array operations. •TensorFlow & Keras: Used to create deep learning models.

Code Structure: The code consists of the following steps;

1. Importing Modules : Introducing the necessary participation into the project.
2. Data Set Upload : Writing the data set consisting of pictures of fish species.
3. Data Set Review : Analysis of the data set.
4. Data Preprocessing : Preprocessing the data setting, making it suitable for training
5. Modeling : Architecture of the model and training of the Model.
6. Model Performance and Evaluation Metrics : Graphical evaluation.
7. Class Predictions on Test Data : Class predictions of the model

Model Structure Main layers used: •Dense (Dense Layer): Fully connected layers. •Flatten: Flattens the data. •Dropout: Prevents overlearning. •BatchNormalization: Normalizes layer outputs during training. Adam optimization algorithm was used.The model uses Sequential Model architecture.

Model Performance Test loss: 0.3259062170982361 Test accuracy: 0.9077777862548828 These results show that the model performs well on the training data and is consistent on the validation set.

Kaggle: https://www.kaggle.com/code/cansuevik/fish-classification-deep-learning