Search
 
loader loading subjects...
cover
Hierarchical Action Selection for Reinforcement Learning in Infinite Mario
Mandar Joshi, Rakesh Khobragade, Saurabh Sarda, Umesh Deshpande, Shiwali Mohan
162 - 167
10.3233/978-1-61499-096-3-162
Frontiers in Artificial Intelligence and Applications
Volume 241: STAIRS 2012
Abstract

In this work, we analyze and improve upon reinforcement learning techniques used to build agents that can learn to play Infinite Mario, an action game. We use the object-oriented representation with the hierarchical RL model as a learning framework. We then extend the idea of hierarchical RL by designing a hierarchy in action selection using domain specific knowledge. Using experimental results, we show that this approach facilitates faster and efficient learning for the domain.

Download
Creative Commons License