Presentation | 2017-07-13 Improving Performance of UCB1-tuned Algorithm by Lebesgue Spectrum Filter Xinyu Cho, Kaori Kuroda, Yukio Murata, Song-Ju Kim, Makoto Naruse, Mikio Hasegawa, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In the researches on asynchronous chaotic CDMA, it has been shown that the sequences, which have negative autocorrelation, minimize asynchronous cross-correlation among the sequences. Such sequences can also improve the solution search performance of combinatorial optimization algorithms. In this report, we introduce such low cross-correlation dynamics to improve the performance of UCB1-tuned which is the best non-parametric algorithm for Multi-armed bandit (MAB) problems. To generate negative autocorrelation dynamics in UCB1-tuned algorithm, we apply the Lebesgue Spectrum Filter (LSF), which has been proposed in a research on the chaotic CDMA. The simulation results show that the performance of UCB1-tuned algorithm can be improved by using the negative autocorrelation dynamics generated by the LSF. We clarify performance improvement by analyzing auto correlation and cross-correlation of the searching sequences. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Multi-armed bandit problem / UCB1-tuned algorithm / Chaotic CDMA / Lebesgue Spectrum Filter |
Paper # | NLP2017-36 |
Date of Issue | 2017-07-06 (NLP) |
Conference Information | |
Committee | NLP |
---|---|
Conference Date | 2017/7/13(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Miyako Island Marine Terminal |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | etc. |
Chair | Masaharu Adachi(Tokyo Denki Univ.) |
Vice Chair | Norikazu Takahashi(Okayama Univ.) |
Secretary | Norikazu Takahashi(Nagaoka Univ. of Tech.) |
Assistant | Toshihiro Tachibana(Shonan Inst. of Tech.) / Masayuki Kimura(Kyoto Univ.) |
Paper Information | |
Registration To | Technical Committee on Nonlinear Problems |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Improving Performance of UCB1-tuned Algorithm by Lebesgue Spectrum Filter |
Sub Title (in English) | |
Keyword(1) | Multi-armed bandit problem |
Keyword(2) | UCB1-tuned algorithm |
Keyword(3) | Chaotic CDMA |
Keyword(4) | Lebesgue Spectrum Filter |
1st Author's Name | Xinyu Cho |
1st Author's Affiliation | Tokyo University of Science(TUS) |
2nd Author's Name | Kaori Kuroda |
2nd Author's Affiliation | Tokyo University of Science(TUS) |
3rd Author's Name | Yukio Murata |
3rd Author's Affiliation | Tokyo University of Science(TUS) |
4th Author's Name | Song-Ju Kim |
4th Author's Affiliation | National Institute for Materials Science(NIMS) |
5th Author's Name | Makoto Naruse |
5th Author's Affiliation | National Institute of Information and Communications Technology(NICT) |
6th Author's Name | Mikio Hasegawa |
6th Author's Affiliation | Tokyo University of Science(TUS) |
Date | 2017-07-13 |
Paper # | NLP2017-36 |
Volume (vol) | vol.117 |
Number (no) | NLP-121 |
Page | pp.pp.47-52(NLP), |
#Pages | 6 |
Date of Issue | 2017-07-06 (NLP) |