Presentation 1996/3/18
Identification and Prediction of Economic Indicators by Answer-in-Weights Network
Takehiko Ogawa, Yukio Kosugi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, we show a network model that covers the two parts of the economic indicators; long-term and periodic components, and that can identify the two parts, simultaneously. This network uses a parametric structure utilizing a priori given information regarding the problem, to improve the generalization ability which cannot be assured by conventional multilayer networks in prediction problems. The model consists of two parts of network arranged in the answer-in-weights structure that requires a solution to appear in the weights and express the fluctuation of economic indicators, by multiplying each output. The simulations on two economic data were done to confirm the model validity.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) economic indicators / generalization ability / multiple network / a priori information
Paper # NC-95-119
Date of Issue

Conference Information
Committee NC
Conference Date 1996/3/18(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) Identification and Prediction of Economic Indicators by Answer-in-Weights Network
Sub Title (in English)
Keyword(1) economic indicators
Keyword(2) generalization ability
Keyword(3) multiple network
Keyword(4) a priori information
1st Author's Name Takehiko Ogawa
1st Author's Affiliation Interdisciplinary Graduate School of Science and Engineering Tokyo Institute of Technology()
2nd Author's Name Yukio Kosugi
2nd Author's Affiliation Interdisciplinary Graduate School of Science and Engineering Tokyo Institute of Technology
Date 1996/3/18
Paper # NC-95-119
Volume (vol) vol.95
Number (no) 598
Page pp.pp.-
#Pages 7
Date of Issue