Presentation | 2013-11-28 Efficient Channel Division and Information Sharing Using Reinforcement Learning for Cooperative Multi-agent Systems Xue ZHANG, Toshiharu SUGAWARA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We propose a method that teams of software agents have been increasingly used in sophisticated real-world systems. In these teams, agents typically need to communicate with each other to cooperate for solving a problem. However, the communication mediums are highly restricted making it infeasible for every agent to broadcast all relevant information to each other. For realizing efficient cooperative actions, not only fair usage of communication bandwidth, but also sharing communication media/channels with collaborating agents is required. To solve this problem, we propose a decentralized algorithm by introducing the two steps of channel division and information sharing with the purpose of maximizing the utility of the whole system. This combines multi-agent learning techniques with fast decision-theoretic reasoning mechanisms to predict the impact that a single agent has on the entire system. We present that our method outperforms the conventional decentralized method by 7.54%, and the results show that our algorithm is more efficient. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Reinforcement learning / Teamwork / Communication / Cooperative agent |
Paper # | AI2013-21 |
Date of Issue |
Conference Information | |
Committee | AI |
---|---|
Conference Date | 2013/11/21(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Artificial Intelligence and Knowledge-Based Processing (AI) |
---|---|
Language | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Efficient Channel Division and Information Sharing Using Reinforcement Learning for Cooperative Multi-agent Systems |
Sub Title (in English) | |
Keyword(1) | Reinforcement learning |
Keyword(2) | Teamwork |
Keyword(3) | Communication |
Keyword(4) | Cooperative agent |
1st Author's Name | Xue ZHANG |
1st Author's Affiliation | School of Fundamental Science and Engineering, Waseda University() |
2nd Author's Name | Toshiharu SUGAWARA |
2nd Author's Affiliation | School of Fundamental Science and Engineering, Waseda University |
Date | 2013-11-28 |
Paper # | AI2013-21 |
Volume (vol) | vol.113 |
Number (no) | 332 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |