/RL-algorithms

Shallow Reinforcement Learning on Atari

Primary LanguagePythonMIT LicenseMIT

Reinforcement Learning algorithms tested in Atari

Implementation of SARSA with Linear Function Approximation using the Basic and B-PROS features described in [1].

Code initially based on the implementation of https://github.com/wulfebw/playing_atari

Backgrounds for Atari games taken from https://github.com/mcmachado/ALEResearch

[1] Yitao Liang, Marlos C. Machado, Erik Talvitie, and Michael Bowling. State of the Art Control of Atari Games Using Shallow Reinforcement Learning