Perceptron Learning Rule Implementation

This repository contains the implementation of the perceptron learning rule for learning a two-input AND gate. The implementation is in Python, using the NumPy library for numerical computations and the Matplotlib library for plotting.

Assignment Details

The assignment is part of the course "Edge and Neuromorphic Computing - CSCE790" at the University of South Carolina for the Spring 2024 semester. The goal is to implement the perceptron learning rule and visualize the decision boundary at the end of each epoch.

Structure

• The report document perceptron.pdf provides a detailed explanation of the assignment, including the problem statement, mathematical formulations, Python code, and results.

• The Python code is implemented in the perceptron.py file. It includes a Perceptron class representing the perceptron, functions for training the perceptron, and plotting decision boundaries.

• Decision boundaries at the end of each epoch are visualized and saved as images (epoch1.png to epoch5.png).

Results

The perceptron learning rule was trained for 5 epochs, and the decision boundaries at the end of each epoch are shown below:

• Epoch 1:
• Epoch 2:
• Epoch 3:
• Epoch 4:
• Epoch 5:

How to Run

To run the code and reproduce the results, follow these steps:

1. Clone the repository: git clone <repository-url>
2. Navigate to the project directory: cd <repository-directory>
3. Run the main Python script: python perceptron.py

Feel free to explore and modify the code for your understanding and learning purposes.