| Presentation | 2017-06-29 A Complex-Valued Reinforcement Learning Method Using Complex-Valued Neural Networks Masaki Mochida, Hidehiro Nakano, Arata Miyauchi, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | This paper proposes the method to approximate the action-value function in complex-valued reinforcement learning by using complex-valued neural networks. Complex-valued reinforcement learning extends action–values to complex numbers and adding action-values to some context information enables the learning algorithm to adapt perceptual aliasing with less memory usage. In order to apply larger environment with the large number of sates, this paper introduces the function approximation for the action-value function by using complex-valued neural networks. We perform simulation experiments on the Mountain Car environment which has continuous states. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | reinforcement learning / perceptual aliasing / complex-valued reinforcement learning / complex-valued neural networks |
| Paper # | CCS2017-1 |
| Date of Issue | 2017-06-22 (CCS) |
| Conference Information | |
| Committee | CCS |
|---|---|
| Conference Date | 2017/6/29(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Ibaraki Univ. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Interaction and Communication, etc. |
| Chair | Naoki Wakamiya(Osaka Univ.) |
| Vice Chair | Mikio Hasegawa(Tokyo Univ. of Science) / Makoto Naruse(NICT) |
| Secretary | Mikio Hasegawa(Osaka Univ.) / Makoto Naruse(Tokyo City Univ.) |
| Assistant | Chisa Takano(Hirishima City Univ.) / Takashi Shimada(Univ. of Tokyo) / Tomoya Suzuki(Ibaraki Univ.) / Ryo Takahashi(AUT) |
| Paper Information | |
| Registration To | Technical Committee on Complex Communication Sciences |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A Complex-Valued Reinforcement Learning Method Using Complex-Valued Neural Networks |
| Sub Title (in English) | |
| Keyword(1) | reinforcement learning |
| Keyword(2) | perceptual aliasing |
| Keyword(3) | complex-valued reinforcement learning |
| Keyword(4) | complex-valued neural networks |
| 1st Author's Name | Masaki Mochida |
| 1st Author's Affiliation | Tokyo City University(Tokyo City Univ.) |
| 2nd Author's Name | Hidehiro Nakano |
| 2nd Author's Affiliation | Tokyo City University(Tokyo City Univ.) |
| 3rd Author's Name | Arata Miyauchi |
| 3rd Author's Affiliation | Tokyo City University(Tokyo City Univ.) |
| Date | 2017-06-29 |
| Paper # | CCS2017-1 |
| Volume (vol) | vol.117 |
| Number (no) | CCS-112 |
| Page | pp.pp.1-5(CCS), |
| #Pages | 5 |
| Date of Issue | 2017-06-22 (CCS) |