/SARSAmora

Reinforcement Learning in Scala

Primary LanguageScala

SARSAmora - Reinforcement Learning library for Scala

SARSAmora is a Scala framework for reinforcement learning. The framework includes abstractions that separate environment interfaces (including OpenAI's Gym using JEP and ScalaPy) and several different state, action and policy representations from a variety of core reinforcement learning algorithms.

The following learning algorithms are currently implemented:

  • SARSA(0)
  • Q-Learning