https://github.com/lambdacasserole/brainy

Neural network implementation in PHP, packaged for Composer.
https://github.com/lambdacasserole/brainy

ai composer neural-network packagist php

Last synced: 5 days ago
JSON representation

Neural network implementation in PHP, packaged for Composer.

• Host: GitHub
• URL: https://github.com/lambdacasserole/brainy
• Owner: lambdacasserole
• License: bsd-3-clause
• Created: 2017-01-30T01:26:43.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2017-03-21T10:30:53.000Z (over 7 years ago)
• Last Synced: 2024-10-10T22:14:56.037Z (27 days ago)
• Topics: ai, composer, neural-network, packagist, php
• Language: PHP
• Homepage:
• Size: 24.4 KB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Brainy

Neural network implementation in PHP, packaged for Composer.

Based heavily on [the Tremani neural network](https://github.com/infostreams/neural-network) by [Edward Akerboom](https://github.com/infostreams) but stripped down, tidied up and packaged for Composer. 

## Installation

Install Brainy via Composer like this:

```bash

composer require lambdacasserole/brainy

```

Or alternatively, if you're using the PHAR (make sure the `php.exe` executable is in your PATH):

```

php composer.phar require lambdacasserole/brainy

```

## Usage

Create a new neural network instance like this:

```php

// Create a new neural network with 3 input neurons, one layer of 4 hidden neurons, and 1 output neuron.

$network = new NeuralNetwork(3, 4, 1);

```

Add training data to your new network thusly:

```php

// Add training data to the network. In this case, we want the network to learn the 'XOR' function.

$network->addTrainingData([-1, -1, 1], [-1]);

$network->addTrainingData([-1, 1, 1], [1]);

$network->addTrainingData([1, -1, 1], [1]);

$network->addTrainingData([1, 1, 1], [-1]);

```

Then begin training:

```php

// Train in a maximum of 1000 epochs to a maximum error rate of 0.01.

$success = $network->train(1000, 0.01);

```

Now put it to work:

```

$output = $network->calculate([-1, -1, 1]); // Gives [-1].

$output = $network->calculate([-1, 1, 1]); // Gives [1].

```

## Compatibility

Uses new array syntax and splats, so won't work on any PHP versions earlier than 5.6.

## Further Reading

The [original repository](https://github.com/infostreams/neural-network) contains more comprehensive documentation, though it may need adjusting slightly due to modifications made to it in this version.