2010 Fiscal Year Final Research Report
A study on exploratory model selection methods for pattern recognition
Project/Area Number |
19500163
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
Perception information processing/Intelligent robotics
|
Research Institution | National Institute of Advanced Industrial Science and Technology |
Principal Investigator |
KURITA Takio National Institute of Advanced Industrial Science and Technology, 大学院・工学研究院, 教授 (10356941)
|
Project Period (FY) |
2007 – 2010
|
Keywords | パターン認識 / 最適化手法 / モデル選択 / 汎化性能 / 物体検出 / 画像認識 / トラッキング |
Research Abstract |
By taking recent developments on optimization algorithms into account, the practical model selection algorithms for pattern recognition such as (1) algorithm to determine the optimum hyper-parameters of the classifier, (2) selection algorithm of the effective features for pattern classification, (3) efficient learning algorithm for object detection in image, (4) automatic design of classifiers with high generalization, (5) feature selection algorithm for object tracking in the image sequences, were proposed and the effectiveness of the optimization algorithms was confirmed.
|
-
-
-
-
[Presentation] パターン認識器の設計のための最適化2011
Author(s)
栗田, 西田、日高, 松川
Organizer
電子情報通信学会パターン認識・メディア理解研究会,信学技報,[特別講演]Vol.110, No.467, PRMU2010-251, pp.85-90
Place of Presentation
産業技術総合研究所,つくば
Year and Date
2011-03-10
-
-
-
-
-
-
-
-
-
-
-
[Presentation] Logistic Discriminant Analysis2009
Author(s)
T.Kurita, K.Watanabe, N.Otsu
Organizer
Proc.of 2009 IEEE International Conference on Systems, Man, and Cybernetics, pp.2236-2241
Place of Presentation
San Antonio, Texas, USA.
Year and Date
20091011-20091014
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-