Study of semi-supervised clustering and challenge to constrained mixture distributions
Project/Area Number |
23500269
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Sensitivity informatics/Soft computing
|
Research Institution | University of Tsukuba |
Principal Investigator |
|
Co-Investigator(Renkei-kenkyūsha) |
ENDO Yasunori 筑波大学, システム情報系, 教授 (10267396)
|
Project Period (FY) |
2011 – 2013
|
Project Status |
Completed (Fiscal Year 2013)
|
Budget Amount *help |
¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2013: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2012: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
Fiscal Year 2011: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
|
Keywords | 半教師付き分類 / 制約クラスタリング / 階層的技法 / 混合分布モデル / COP K-means / ファジィクラスタリング / インダクティブクラスタリング / クラスタリング / 拡張K-means / クラスター評価 / 2段階クラスタリング / 非対称データ / 階層的クラスタリング / 制約K-means / 制約混合分布 / 非対称類似度 / ファジィ近傍モデル / Twitter / 正定値カーネル |
Research Abstract |
Existing methods of semi-supervised clustering and proposed methods by the authors are compared using artificial and repository data. Main results are as follows. 1. Agglomerative clustering performs as well as mixture distribution models in constrained clustering. 2.Methods of fuzzy clustering generalize basic mixture distribution models for semi-supervised classification. 3. Extensions of COP K-means have been proposed but they did not perform as well as constrained mixture distribution method. 4. The concept of inductive clustering has been proposed and its methodological usefulness has been shown. 5. Real data using Twitter have been analyzed using semi-supervised clustering and its effects have been investigated.
|
Report
(4 results)
Research Products
(60 results)
-
-
-
-
-
[Journal Article]2013
Author(s)
S. Miyamoto, S. Takumi
-
Journal Title
Journal of Advanced Computational Intelligence and Intelligent Informatics
Volume: 17
Pages: 504-510
Related Report
Peer Reviewed
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-