$ Rust Neural Network
This project implements a basic feedforward neural network in Rust. It's designed to be a clear and understandable example of how neural networks work, incorporating fundamental concepts like activation functions and backpropagation for training.
### features
- - Activation Functions — Supports ReLU, Sigmoid, Tanh, and Linear activation functions.
- - Feedforward Propagation — Calculates the output of the network given an input.
- - Backpropagation Algorithm — Implements backpropagation for training the network.
- - Weight Initialization — Uses He or Xavier scaling for initializing weights.
- - Serialization — Network structure can be serialized and deserialized using `serde`.
### example usage
The `main.rs` example demonstrates how to create, train, and use the neural network for a simple task.
It's a foundational project for anyone looking to understand the mechanics of neural networks from a low-level perspective using Rust.