Presentation	1998/7/27 A Data Selection and Training Method for Generalization Kazuyuki HARA, Kenji NAKAYAMA,
PDF Download Page	PDF download Page Link
Abstract(in Japanese)	(See Japanese page)
Abstract(in English)	In this paper, a training data selection method for multilayer neural networks which guarantees generalization performance is proposed. A pairing method selects the nearest neighbor data by finding the nearest data in the different classes, and is used to select the data which gurantee generalization performance. For the training with selected data, we propose the sigmoid function switching method. This method starts with unipolar sigmoid function, and then it swtiches to the bioplar along the training process. Effciency of these methods are evaluated by computer simulation.
Keyword(in Japanese)	(See Japanese page)
Keyword(in English)	Multilayer neural network / Class boundary / Sigmoid function Switching
Paper #	NC98-35
Date of Issue

Conference Information
Committee	NC
Conference Date	1998/7/27(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 Data Selection and Training Method for Generalization
Sub Title (in English)
Keyword(1)	Multilayer neural network
Keyword(2)	Class boundary
Keyword(3)	Sigmoid function Switching
1st Author's Name	Kazuyuki HARA
1st Author's Affiliation	Dept. Elec. & Info. Eng., Tokyo Metropolitan College of Technology()
2nd Author's Name	Kenji NAKAYAMA
2nd Author's Affiliation	Dept. Elec. & Comp. Eng., Kanazawa University
Date	1998/7/27
Paper #	NC98-35
Volume (vol)	vol.98
Number (no)	219
Page	pp.pp.-
#Pages	8
Date of Issue