| 大会名称 |
| 2019年 ソサイエティ大会 |
| 大会コ－ド |
| 2019S |
| 開催年 |
|
2019
|
| 発行日 |
|
2019/8/27
|
| セッション番号 |
|
BS-2
|
| セッション名 |
|
Beyond 5G / 6Gを目指した無線ネットワークの進化
|
| 講演日 |
|
2019/9/11
|
| 講演場所(会議室等) |
|
C棟 4F Ｃ402講義室
|
| 講演番号 |
|
BS-2-6
|
| タイトル |
|
Source Separation Learning in epsilon-Vanishing Polynomial Network
|
| 著者名 |
|
○LU WANG, Tomoaki Ohtsuki,
|
| キーワード |
|
Underdetermined BSS, vanishing polynomial networks, nonlinear mixture, sparse coding, time-frequency representation
|
| 抄録 |
|
Similar to the deep architectures, a novel polynomial network is presented to extract the nonlinearity caused by mixing function. Our network begins with the polynomial of degree $1$, up to build an output layer that can represent data with a small bias by a good approximate basis. Relying on several transformations of the input data, with higher-level representation from lower-level ones, the networks are to fulfill a mapping implicitly to the high-dimensional space. Once the polynomial networks are built, the coefficient matrix can be estimated by solving an optimization problem on the coding coefficient vector.
|
| 本文pdf |
|
PDF download
|
