Presentation 1999/3/18
A Proposal of an Improved Function Synthesis Algorithm with Genetic Programming
TETSUYA SEKIOKA, NOBUO FUNABIKI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The function synthesis problem for an unknown system requires to produce an approximate function which satisfies the relationships of given input and output data sets of the system. This problem is essential for the proper control of the system. When the system has the non-linearity characteristics, this problem becomes very difficult where the number of combinations of variables, constants, and operators quickly increases the computational complexity. This paper presents three improved methods for the genetic programming algorithm which is one of the function synthesis methods. They are the gradual expansion of function trees, the constant optimization by a local search, and the partial structure fixation of function trees based on the characteristics of the target system. The performance evaluation through the two-spirals problem and the sunspot prediction problem shows the effectiveness of the proposed three methods.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Function synthesis / Genetic Programming / two-spirals / sunspots
Paper # NLP98-123
Date of Issue

Conference Information
Committee NLP
Conference Date 1999/3/18(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 Nonlinear Problems (NLP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Proposal of an Improved Function Synthesis Algorithm with Genetic Programming
Sub Title (in English)
Keyword(1) Function synthesis
Keyword(2) Genetic Programming
Keyword(3) two-spirals
Keyword(4) sunspots
1st Author's Name TETSUYA SEKIOKA
1st Author's Affiliation Department of Informatics and Mathematical Science, Graduate School of Engineering Science, Osaka University()
2nd Author's Name NOBUO FUNABIKI
2nd Author's Affiliation Department of Informatics and Mathematical Science, Graduate School of Engineering Science, Osaka University
Date 1999/3/18
Paper # NLP98-123
Volume (vol) vol.98
Number (no) 663
Page pp.pp.-
#Pages 8
Date of Issue