講演抄録/キーワード |
講演名 |
2009-03-13 15:50
Semi-supervised learning scheme using Dirichlet process EM-algorithm ○Tomoaki Kimura・Yohei Nakada(Waseda Univ.)・Arnaud Doucet(ISM)・Takashi Matsumoto(Waseda Univ.) PRMU2008-251 |
抄録 |
(和) |
(まだ登録されていません) |
(英) |
Learning with dataset which contains both labeled data and unlabeled data
is often called semi-supervised learning problem.
In the last decade,
semi-supervised learning problem has become one of the important
research problems in many fields.
This article presents a novel semi-supervised learning scheme
using Bayesian Maximum A Posteriori Expectation Maximization (MAP-EM)
algorithm with Dirichlet process prior (stick-breaking
representation).
The proposed scheme enables us to estimate mixture model under unknown
number of components
and provides a simpler implementation than other implementations such
as Markov Chain Monte Carlo (MCMC) implementations.
Two examples of Gaussian mixture are examined to validate the proposed scheme. |
キーワード |
(和) |
/ / / / / / / |
(英) |
learning system / semi-supervised learning / Dirichlet process / stick breaking / EM / Gaussian mixture / / |
文献情報 |
信学技報, vol. 108, no. 484, PRMU2008-251, pp. 77-82, 2009年3月. |
資料番号 |
PRMU2008-251 |
発行日 |
2009-03-06 (PRMU) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
PRMU2008-251 |