Home

Project description

This repository features a set of tools/applications designed to make experiments with Reinforcement Learning algorithms on control tasks with continuous state and action spaces. Experiments are designed using a GUI to make things easier to end-users.

Main applications:

• RLSimion: a C++ console application that does the actual Reinforcement Learning, taking an XML file as input and generating a log as output.
• SimionSrcParser: a C# program that parses the C++ code of a Simion app (only RLSimion, so far) and exports editable parameters to /config/definitions.xml
• Badger: a C# WPF GUI tool that allows to design experiments and run them locally or in remote machines.
• Herd agent: a C# Windows service that needs to be installed in those machines intended to work as slaves from Badger.

The folder structure is described here

The basic programming rules are given programming rules.

Here you can find how to set directories in new projects and some Visual Studio 2015-related stuff. A mini-tutorial on how to use GitHub from Visual Studio 2015 here.

Acknowledgements

The code features contributions from:

• Unai Tercero (Badger and Herd Agent)
• Asier Rodríguez (Bullet worlds)
• Alejandro Guerra (Badger and Herd Agent)
• Roland Zimmermann (OffPAC, INAC, ... and all about CNTK and Deep RL) Other than that, except a few fixes by others (check the commmits), the rest of the code has been written and is maintained by Borja Fernández-Gauna from the University of the Basque Country (UPV/EHU), so any question related to it can be directed to my email address: borja.fernandez'at'ehu.eus.