| Presentation | 2018-12-23 On Reward Sharing for Multi-agent Games Using Deep Reinforcement Learning Kei Watanabe, Toshihiro Tachibana, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this paper, we consider a method to realize multi agent system with multiple agents cooperating by deep reinforcement learning. Furthermore, we conduct simulation experiments. Here we use the game engine "Unity" published by Unity Technologies and the machine learning framework "UnityML-Agents". We will use it to build a learning environment and create an agent architecture. The target environment here is a game where multiple agents keep points. We experimentally verified the learning method of the inference model which manipulates the agent using deep reinforcement learning, the relationship of how to give the reward and the behavior. In this paper, we report on the results. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Deep reinforcement learning / Multi-agents / Reward / Unity ML-Agents / TensorFlow |
| Paper # | CAS2018-107,ICD2018-91,CPSY2018-73 |
| Date of Issue | 2018-12-14 (CAS, ICD, CPSY) |
| Conference Information | |
| Committee | ICD / CPSY / CAS |
|---|---|
| Conference Date | 2018/12/21(3days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Hideto Hidaka(Renesas) / Koji Nakano(Hiroshima Univ.) / Hideaki Okazaki(Shonan Inst. of Tech.) |
| Vice Chair | Makoto Nagata(Kobe Univ.) / Hidetsugu Irie(Univ. of Tokyo) / Takashi Miyoshi(Fujitsu) / Taizo Yamawaki(Hitachi) |
| Secretary | Makoto Nagata(Panasonic) / Hidetsugu Irie(Tohoku Univ.) / Takashi Miyoshi(Utsunomiya Univ.) / Taizo Yamawaki(Hokkaido Univ.) |
| Assistant | Hiroyuki Ito(Tokyo Inst. of Tech.) / Masatoshi Tsuge(Socionext) / Tetsuya Hirose(Kobe Univ.) / Yasuaki Ito(Hiroshima Univ.) / Tomoaki Tsumura(Nagoya Inst. of Tech.) / Motoi Yamaguchi(Renesas Electronics) |
| Paper Information | |
| Registration To | Technical Committee on Integrated Circuits and Devices / Technical Committee on Computer Systems / Technical Committee on Circuits and Systems |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | On Reward Sharing for Multi-agent Games Using Deep Reinforcement Learning |
| Sub Title (in English) | |
| Keyword(1) | Deep reinforcement learning |
| Keyword(2) | Multi-agents |
| Keyword(3) | Reward |
| Keyword(4) | Unity ML-Agents |
| Keyword(5) | TensorFlow |
| 1st Author's Name | Kei Watanabe |
| 1st Author's Affiliation | Shonan Institute of Technology(Shonan Isnt. of Tech.) |
| 2nd Author's Name | Toshihiro Tachibana |
| 2nd Author's Affiliation | Shonan Institute of Technology(Shonan Isnt. of Tech.) |
| Date | 2018-12-23 |
| Paper # | CAS2018-107,ICD2018-91,CPSY2018-73 |
| Volume (vol) | vol.118 |
| Number (no) | CAS-373,ICD-374,CPSY-375 |
| Page | pp.pp.109-114(CAS), pp.109-114(ICD), pp.109-114(CPSY), |
| #Pages | 6 |
| Date of Issue | 2018-12-14 (CAS, ICD, CPSY) |