Skip to content
A workbench for online model-free Reinforcement Learning on continuous control problems
C++ C# Shell HTML C F*
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.github/ISSUE_TEMPLATE
3rd-party
Badger
BadgerConsole
Debug
RLSimion
bin
config
experiments/examples
installers
tests
tools
.gitattributes
.gitignore
CODE_OF_CONDUCT.md
CONTRIBUTING.md
LICENSE
README.md
SimionZoo.sln
azure-pipelines.yml
linux-ci-build.sh

README.md

SimionZoo

Branch Build status
Master Build Status
Develop Build Status

About the project

This project is mantained by the Group of Computational Intelligence (University of the Basque Country - UPV/EHU): http://www.ehu.eus/ccwintco

Our goal is to build a Reinforcement Learning experimenting environment in which the user can conduct experiments using a GUI without needing to code, install any dependecies, or run a configuration process.

The focus of the project is set on solving control problems with continuous action and states using Reinforcement Learning methods. The main features of the project can be summarized:

  • Parameter sweeps using 'forks' on the parameters
  • Real-time and off-line experiment visualization
  • Built-in result analysis tools
  • Distributed execution of experiments (distribution is done via the 'forks')

Getting started

In the wiki you can find information about how to get started with this project, either as an user or as a developer.

Acknowledgements

This project has been mainly contributed by Unai Tercero, Asier Rodriguez, Roland Zimmermann, José Alejandro Guerra Denis and Borja Fernández Gauna.

You can’t perform that action at this time.