Title:
【発明の名称】ニューラルネットワークシステム
Document Type and Number:
Japanese Patent JP2860057
Kind Code:
B2
Abstract:
The partial linear regression model structure in which data generated in connection with the time flowing is processed to find a function relationship of the system from which the data is generated, comprises a self partitioning network(10) for partitioning input space into a plurality of local areas which are not overlapped, and a feed forward mapping network(20) including a hidden layer and an output layer, for obtaining linear mapping for each local area in response to the output of the self partitioning network(10).
Inventors:
SAI CHINEI
KICHI RIBAN
KICHI RIBAN
Application Number:
JP18371794A
Publication Date:
February 24, 1999
Filing Date:
August 04, 1994
Export Citation:
Assignee:
KANKOKU DENSHI TSUSHIN KENKYUIN
International Classes:
G06G7/60; G06F15/18; G06N3/04; (IPC1-7): G06F15/18; G06G7/60
Domestic Patent References:
| JP4155480A |
Other References:
【文献】Jin Young Choi,Rhee Man Kil,Chong-Ho Choi,”Piecewise Linear Regression Networks and Its Application to Time Series Prediction”,Proceedings of 1993 International Joint Conference on Neural Networks,vol.2,p.1349-p.1352(1993.10)
Attorney, Agent or Firm:
Kazuko Tomita (1 outside)