LiteNN

A Lite ,easy for learning neural networks (MLP) implementation project.

About

The project has implemented as follow:

• Basic matrix calculation implements.
• Backpropagation with matrix form.
• A Perceptron with sigmoid activate function.
• A Three-layered MLP with sigmoid activate function.
• Demo training for the XOR problem.

Getting Started & Run

• Linux user:

$ sudo apt-get install git-all
$ sudo apt-get install build-essential
$ git clone https://github.com/whuang022ai/LiteNN.git
$ cd LiteNN/LiteNN
$ make -f Makefile.linux
$ ./LiteNN

• Mac OS X user (Please install brew before you typing the following commands.):

$ brew install git
$ brew install gcc
$ brew install make
$ git clone https://github.com/whuang022ai/LiteNN.git
$ cd LiteNN/LiteNN
$ make -f Makefile.mac
$ ./LiteNN

• Windows user:

Install git,g++,make for Windows as following and seeting the System Path:

• Windows Git

• Windows MinGW(g++)

• Windows make

$ git clone https://github.com/whuang022ai/LiteNN.git
$ cd LiteNN/LiteNN
$ make -f Makefile.win
$ LiteNN.exe

Project Usage

This project is a demo code for classes as following:

• AITLK 2018

• AITLK 2019