| 大会名称 |
| 2016年 ソサイエティ大会 |
| 大会コ－ド |
| 2016S |
| 開催年 |
|
2016
|
| 発行日 |
|
2016-09-06
|
| セッション番号 |
|
A-10
|
| セッション名 |
|
システム数理と応用
|
| 講演日 |
|
2016/9/20
|
| 講演場所(会議室等) |
|
工学部 情報科学研究科棟 A22
|
| 講演番号 |
|
A-10-1
|
| タイトル |
|
Stochastic low-rank tensor completion: a Riemannian manifold preconditioning approach
|
| 著者名 |
|
○Hiroyuki Kasai, Bamdev Mishra,
|
| キーワード |
|
manifold optimzation, tensor completion
|
| 抄録 |
|
We propose a novel Riemannian manifold preconditioning approach for the stochastic tensor completion problem with rank constraint. A novel Riemannian metric or inner product is proposed that exploits the least-squares structure of the cost function and takes into account the structured symmetry that exists in Tucker decomposition. The specific metric allows to use the versatile framework of Riemannian optimization on quotient manifolds to develop preconditioned stochastic gradient descent algorithms for online setups.
|
| 本文pdf |
|
PDF download
|
