Neural Network From Scratch

Overview

This repository is a learning project where I explore how neural networks are built from scratch using Java and Python. The goal is to understand the inner workings of neural networks without relying on high-level machine learning libraries.

Features

• Implementation of basic neural network components.
• Support for multiple activation functions.
• Training and testing using custom datasets.
• Implementations in both Java and Python.

Setup

Python

1. Clone the repository:

   git clone https://github.com/sky0walker99/NNFromScratch.git
   cd NNFromScratch

2. Run a sample script:

   python src/main.py

Java

1. Ensure you have JDK installed.
2. Navigate to the Java source directory:

   cd src/java

3. Compile and run:

   javac NeuralNetwork.java
   java NeuralNetwork

Learning Objectives

• Understanding fundamental concepts of neural networks.
• Implementing forward and backward propagation.
• Exploring different optimization techniques.
• Comparing implementations in Python and Java.

Contributing

This repository is for personal learning, but contributions and discussions are welcome!

License

This project is licensed under the MIT License.