Skip to content

Latest commit

 

History

History
28 lines (21 loc) · 770 Bytes

README.md

File metadata and controls

28 lines (21 loc) · 770 Bytes
Raw

Artificial Neural Network and reinforcement learning (Q-Learning) Blinky or Hello World C-Project

This is an easy example for the use of fann as ANN lib and an Q-Learning algorithm to train a network

Hardware

STM32F4 Discovery Board -> CortexM4 with FPU

solder a extra button from GND to PC6!

License

GPLv2

Libs

fann (LGPL see http://leenissen.dk/) (added as static library to this project)

Software

System Workbench (Eclipse based) STM CubeMX

Introduction

Press the blue Discovery button if you wan't to train the state change. Press the new button to "untrain" this action. Do this a 100 times and the pattern will be displayed after training.