JSAI Technical Report, Type 2 SIG
Online ISSN : 2436-5556
A Preliminary Study of a Table Compression Method Using a Unification Mechanism for a Hierarchical Reinforcement Learning
Yuuji ICHISUGINaoto TAKAHASHIHidemoto NAKADATakashi SANO
Author information
RESEARCH REPORT / TECHNICAL REPORT FREE ACCESS

2018 Volume 2018 Issue AGI-010 Pages 02-

Details
Abstract

We propose a method of compressing the size of the table of the action value function for the hierarchical reinforcement learning architecture RGoal, using a mechanism of unification. Moreover, we study the characteristics of RGoal as a program synthesis system.

Content from these authors
© 2018 Authors
Previous article Next article
feedback
Top