| Presentation | 2018-03-06 Selecting discriminative and representative patterns from sequence data: an approach based on classification model and morse complex Masayuki Karasuyama, Ichiro Takeuchi, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | We study classification problem on the sequences of continuous observations. In particular, we are interested in identifying class specific consecutive subsequences. We consider a model which only has local dependency in the sequence, and define a class specificity score of a given consecutive sub-sequence. By partitioning the input sub-sequence space based on morse complex of the class specificity score function, representatives of class specific patterns can be identified. A remarkable property of our approach is that all procedures can be performed without depending any initialization, and the optimal values can be obtained uniquely. Numerical experiments show effectiveness of our approach. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | time-series classification / morse complex / latent variable model |
| Paper # | IBISML2017-101 |
| Date of Issue | 2018-02-26 (IBISML) |
| Conference Information | |
| Committee | IBISML |
|---|---|
| Conference Date | 2018/3/5(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Nishijin Plaza, Kyushu University |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Statisitical Mathematics, Machine Learning, Data Mining, etc. |
| Chair | Kenji Fukumizu(ISM) |
| Vice Chair | Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Kyoto Univ.) |
| Secretary | Masashi Sugiyama(Nagoya Inst. of Tech.) / Hisashi Kashima(Univ. of Tokyo) |
| Assistant | Tomoharu Iwata(NTT) / Toshihiro Kamishima(AIST) |
| Paper Information | |
| Registration To | Technical Committee on Infomation-Based Induction Sciences and Machine Learning |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Selecting discriminative and representative patterns from sequence data: an approach based on classification model and morse complex |
| Sub Title (in English) | |
| Keyword(1) | time-series classification |
| Keyword(2) | morse complex |
| Keyword(3) | latent variable model |
| 1st Author's Name | Masayuki Karasuyama |
| 1st Author's Affiliation | Nagoya Institute of Technology/National Institute for Material Science/Japan Science and Technology Agency(Nagoya Inst. of Tech./NIMS/JST) |
| 2nd Author's Name | Ichiro Takeuchi |
| 2nd Author's Affiliation | Nagoya Institute of Technology/National Institute for Material Science/Institute of Physical and Chemical Research(Nagoya Inst. of Tech./NIMS/RIKEN) |
| Date | 2018-03-06 |
| Paper # | IBISML2017-101 |
| Volume (vol) | vol.117 |
| Number (no) | IBISML-475 |
| Page | pp.pp.77-84(IBISML), |
| #Pages | 8 |
| Date of Issue | 2018-02-26 (IBISML) |