Presentation 2005/3/23
A Quick Learning method with Gamble
Yohei TADEUCHI, Koichiro YAMAUCHI, Takashi OMORI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this study we try to construct a real time model selection method for machine learning systems. To achieve this, the system has to select appropriate model by observations of a small number of instances. However, such the model selection is not guaranteed to be achieved successfully. Therefore, we think that the system must use prior knowledge and select the model with gamble. So we investigate the problem solving strategy of human being and propose the machine learning system inspired from the findings. In the method, at first, the system stores the networks which have completed the learning. Then, the system creates solution candidates for current problem using the stored networks. And the system tries to complete the learning quickly by applying the best candidate tentatively.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Radial Basis Function / Minimal Resource Allocating Network(MRAN) / Quick Learning / Model-selection
Paper # NC2004-218
Date of Issue

Conference Information
Committee NC
Conference Date 2005/3/23(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Quick Learning method with Gamble
Sub Title (in English)
Keyword(1) Radial Basis Function
Keyword(2) Minimal Resource Allocating Network(MRAN)
Keyword(3) Quick Learning
Keyword(4) Model-selection
1st Author's Name Yohei TADEUCHI
1st Author's Affiliation Formative System Engineering Laboratory, Resarch Group of Complex Systems Engineering, Hokkaido University()
2nd Author's Name Koichiro YAMAUCHI
2nd Author's Affiliation Formative System Engineering Laboratory, Resarch Group of Complex Systems Engineering, Hokkaido University
3rd Author's Name Takashi OMORI
3rd Author's Affiliation Formative System Engineering Laboratory, Resarch Group of Complex Systems Engineering, Hokkaido University
Date 2005/3/23
Paper # NC2004-218
Volume (vol) vol.104
Number (no) 760
Page pp.pp.-
#Pages 6
Date of Issue