scispace - formally typeset
Search or ask a question
JournalISSN: 1883-8014

Journal of Advanced Computational Intelligence and Intelligent Informatics 

Fuji Technology Press Ltd.
About: Journal of Advanced Computational Intelligence and Intelligent Informatics is an academic journal published by Fuji Technology Press Ltd.. The journal publishes majorly in the area(s): Fuzzy logic & Computer science. It has an ISSN identifier of 1883-8014. It is also open access. Over the lifetime, 2348 publications have been published receiving 10904 citations.


Papers
More filters
Journal ArticleDOI
TL;DR: Opposition-based reinforcement learning, inspired by opposition-based learning, is introduced, to speed up convergence by Considering opposite actions simultaneously enables individual states to be updated more than once shortening exploration and expediting convergence.
Abstract: Reinforcement learning is a machine intelligence scheme for learning in highly dynamic, probabilistic environments By interaction with the environment, reinforcement agents learn optimal control policies, especially in the absence of a priori knowledge and/or a sufficiently large amount of training data Despite its advantages, however, reinforcement learning suffers from a major drawback – high calculation cost because convergence to an optimal solution usually requires that all states be visited frequently to ensure that policy is reliable This is not always possible, however, due to the complex, high-dimensional state space in many applications This paper introduces opposition-based reinforcement learning, inspired by opposition-based learning, to speed up convergence Considering opposite actions simultaneously enables individual states to be updated more than once shortening exploration and expediting convergence Three versions of Q-learning algorithm will be given as examples Experimental results for the grid world problem of different sizes demonstrate the superior performance of the proposed approach

188 citations

Performance
Metrics
No. of papers from the Journal in previous years
YearPapers
202384
2022113
202183
2020101
2019132
2018111