PCA and Feature Extraction Project

This repository contains Python implementations of various deep learning algorithms for classification, clustering, dimensionality reduction, and feature extraction.

Overview

The project includes implementations of the following algorithms:

1. Principal Component Analysis (PCA):

   • PCA is implemented for dimensionality reduction and feature extraction.
   • The algorithm computes the principal components and reconstructs the data based on the selected components.

2. Adaptive Resonance Theory 1 (ART1) Network:

   • The ART1 network is implemented for clustering unlabeled data.
   • It trains the network on the Iris dataset and predicts clusters.
   • The repository also includes visualization of clusters and prototypes in bottom-up and top-down space.

Contents

• PCA:

  • Implementation of PCA algorithm.
  • Visualization of transformed data in a scatter plot.
  • Computation of covariance matrix.

• ART1 Network:

  • Implementation of ART1 network for clustering.
  • Training the network on the Iris dataset.
  • Visualization of clusters and prototypes.

Getting Started

Prerequisites

Make sure you have the following dependencies installed:

• NumPy
• Matplotlib
• scikit-learn

Usage

1. Clone the repository:

   git clone <repository_url>

2. Navigate to the repository directory:

   cd <repository_directory>

3. Run the Python scripts for the algorithms you want to use:

   • For PCA: python pca.py
   • For ART1 Network: python art1_network.py

4. View the output and visualizations provided by the scripts.

Author

• Mohamed Abobakr
• Nour Hamdy
• Mariam Elsabagh