This crate is a modular implementation of a neural network. It was created with the intention to make the implementation of any learning method for the neural network really simple and without a runtime cost. If you want to know how to implement any learning technique you can take a look at the src/back_prop folder where you can find an implementation of back propagation. If you're just interested in using a neural network to solve problems you can follow the tutorial down below on how to add it to your project and how to use it.
Add neural_network to your project
Add this line to your dependencies in your Cargo.toml file
neural_network = "0.1.3"
Then, add this at the beginning of your src/main.rs or src/lib.rs file
extern crate neural_network;
And you're good to go!
Examples of using back propagation
Here's an example of how to use back propagation to train a neural network to return the sin and the cos of a number. You'll also see how to save the network to a file and reload it.
use *; use ; use rand;