Main Page: Difference between revisions
From Artificial Neural Network for PHP
No edit summary |
No edit summary |
||
| Line 3: | Line 3: | ||
This project realizes a '''neural network topology called ''multilayer perceptron'' for PHP 5.x''' environments. The source code is based on a work by ''Eddy Young'' in 2002. Several improvements and changes on this implementation are done by ''Thomas Wien'' in 2007. |
This project realizes a '''neural network topology called ''multilayer perceptron'' for PHP 5.x''' environments. The source code is based on a work by ''Eddy Young'' in 2002. Several improvements and changes on this implementation are done by ''Thomas Wien'' in 2007. |
||
== Versions and Change-Log == |
|||
---- |
|||
'''Version 1.0 by Eddy Young''' (2002) |
'''Version 1.0 by Eddy Young''' (2002) |
||
| Line 25: | Line 25: | ||
* Fixing bug: initializing inputs to all hidden layers |
* Fixing bug: initializing inputs to all hidden layers |
||
* Fixing bug: training for first hidden layer was skipped |
* Fixing bug: training for first hidden layer was skipped |
||
== Todo == |
|||
* Separation of classes to several files |
|||
* Graphical output of neural network values |
|||
== Overview == |
== Overview == |
||
Revision as of 20:44, 17 December 2007
ANN - Artificial Neural Network for PHP 5.x
This project realizes a neural network topology called multilayer perceptron for PHP 5.x environments. The source code is based on a work by Eddy Young in 2002. Several improvements and changes on this implementation are done by Thomas Wien in 2007.
Versions and Change-Log
Version 1.0 by Eddy Young (2002)
- Initial version
Version 2.0 by Thomas Wien (17.12.2007)
- PHP 5.x support
- PHPDoc documentation
- Momentum support
- Linear / binary output
- ANN_InputValue + ANN_OutputValue classes
- Exceptions
- Threshold function
- Tangens hyperbolicus transfer function
- Several performance issues
- Avoiding array_keys & srand
- Changes in saving and loading network
- Printing network details to browser
- Fixing bug: initializing inputs to all hidden layers
- Fixing bug: training for first hidden layer was skipped
Todo
- Separation of classes to several files
- Graphical output of neural network values
