Azamat-Urinboyev / cats-vs-dogs

Cats vs Dogs Image Classification

This project involves building a machine learning model to classify images of cats and dogs using a convolutional neural network (CNN). The dataset used for training and testing the model is from the Kaggle Dogs vs. Cats competition.

Project Structure

• cats_vs_dogs.ipynb: Jupyter notebook containing the code for data loading, preprocessing, model building, training, and evaluation.
• dogs-vs-cats/: Directory containing the dataset files.

Requirements

• Python 3.x
• Libraries:
  • numpy
  • pandas
  • keras
  • sklearn
  • matplotlib
  • google.colab (for file uploads in Google Colab)

You can install the required libraries using pip:

pip install numpy pandas keras scikit-learn     matplotlib

Dataset

The dataset can be downloaded from the Kaggle Dogs vs. Cats competition. Make sure to place the dataset files in the dogs-vs-cats/ directory.

Usage

1. Clone the repository:

    git clone https://github.com/yourusername/cats_vs_dogs.git
    cd cats_vs_dogs

2. Run the Jupyter Notebook:

    jupyter notebook cats_vs_dogs.ipynb

3. Follow the steps in the notebook:

• Import necessary libraries.
• Upload and prepare the dataset.
• Define image properties.
• Build and train the CNN model.
• Evaluate the model performance.

Results

The model achieves an accuracy of approximately 98% on the test set.

Acknowledgments

Kaggle for providing the dataset.

Google Colab for providing a free and convenient platform for running Jupyter notebooks.