Skip to content

Latest commit

 

History

History
43 lines (31 loc) · 2.62 KB

README.md

File metadata and controls

43 lines (31 loc) · 2.62 KB

Adaline and Madaline Network for Classification 🔍🧠

Python Machine Learning

This repository contains implementations of the Adaline (Adaptive Linear Neuron) and Madaline (Multiple Adaptive Linear Neuron) networks, showcasing foundational neural network concepts through binary classification tasks.

Features 🌟

  • Adaline network implementation for binary classification with visualization of decision boundaries.
  • Madaline network implementation with multiple neurons for enhanced pattern distinction.
  • Demonstrates the use of different activation functions and learning rates.
  • Includes detailed data generation, training processes, and visualization of model training and classification results.

Setup and Installation 🛠️

  1. Clone the repository.
  2. Install the required Python libraries specified in requirements.txt.

Adaline Network 📈

  • Adaline is a simple type of single-layer neural network with weights adjusted according to the difference between the actual and predicted outputs (delta rule).
  • The code provides functions to initialize the network, perform forward propagation, and apply weight updates.

Madaline Network 📊

  • Madaline extends the Adaline by introducing multiple layers of neurons, allowing for more complex decision boundary formation.
  • The code includes the mechanism for Madaline's forward propagation, decision making, and weight updates based on the training data.

Usage 🚀

  • To run the Adaline network: execute the Adaline training function with the desired dataset and hyperparameters.
  • For the Madaline network: initialize and run the Madaline training function, specifying the network architecture and training parameters.

Results and Evaluation 📊

  • Both implementations output graphs showing the loss over training epochs, decision boundaries, and the classification's confusion matrix.
  • The effectiveness of each model is quantified through performance metrics like accuracy and the mean squared error.

Contributing 🤝

Contributions, bug reports, and feature requests are welcome! Feel free to fork the repository, make your changes, and submit a pull request.

License 📜

This project is licensed under the MIT License - see the LICENSE file for more details.

Acknowledgements 🙌

  • The AI and neural network community for foundational theories and practices that inspire these implementations.

For more information and to view the source code, visit the GitHub repository.