Skip to content

Simple, adaptable, fully connected neural network with built-in image support

License

Notifications You must be signed in to change notification settings

devNegative-asm/zJavaAI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 

Repository files navigation

zJavaAI

Simple, adaptable, fully connected neural network with built-in image support.
Compile with mtj-1.0 and Jama 1.0.3.

Usage:

  1. Enter Interface.java, change FILE_NAME_STRING to the directory holding your training data.
  2. Set AI_NAME to the name of the file storing the neural net. this file can get large.
  3. Set Bx and By to the desired image size, and MONOCHROME to true or false
  4. set TRAIN to true while training. then to false to simply view generated images.
  5. Set up other metadata as necessary or desired.

The neural net can function with other data types other than images, but you first need functions that convert your data to and from nVector's.
The NeuralNetHelper class contains a straightforward framework to construct a neural network for any data type.
When using the neural net on non-image data types, it is recommended to not use the Interface class.

About

Simple, adaptable, fully connected neural network with built-in image support

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages