| Presentation | 2012-01-26 The learning theory and algorithm of latent multi-dynamical systems : Implementation by higher-order topographic mapping Tetsuo FURUKAWA, Takashi OHKUBO, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | The purpose of this paper is to establish the learning theory of multiple dynamical systems, as well as to develop the algorithm. To achieve this, we have been studied with setting three agendas. The first agenda is to establish the theoretical definition of the task. The second agenda is to make clear what kind of algorithm is required to solve the task. And the third one is to develop such an algorithm. With respect to the first point, we have found a definion that the multi-dynamical system learning is the estimation task of a system space which consists of a group of dynamical systems with latent state variables. By applying this definition, the algorithm should have the ability of estimating a nonlinear product manifold from observed datasets. Thus the nonlinear tensor decomposition is required. Finally, we have developed an algorithm of the multi-dynamical system learning by making a product space generated by a set of self-organizing maps. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | latent dyanmics / multi-system learning / product manifold / tensor / topographic mapping / self-organizing map / transfer learning |
| Paper # | NC2011-107 |
| Date of Issue |
| Conference Information | |
| Committee | NC |
|---|---|
| Conference Date | 2012/1/19(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 | Neurocomputing (NC) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | The learning theory and algorithm of latent multi-dynamical systems : Implementation by higher-order topographic mapping |
| Sub Title (in English) | |
| Keyword(1) | latent dyanmics |
| Keyword(2) | multi-system learning |
| Keyword(3) | product manifold |
| Keyword(4) | tensor |
| Keyword(5) | topographic mapping |
| Keyword(6) | self-organizing map |
| Keyword(7) | transfer learning |
| 1st Author's Name | Tetsuo FURUKAWA |
| 1st Author's Affiliation | Department of Brain Science and Engineering Kyushu Institute of Technology() |
| 2nd Author's Name | Takashi OHKUBO |
| 2nd Author's Affiliation | Department of Brain Science and Engineering Kyushu Institute of Technology |
| Date | 2012-01-26 |
| Paper # | NC2011-107 |
| Volume (vol) | vol.111 |
| Number (no) | 419 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |