Author(s): Huang, CY (Huang, Chien-Yu); Chen, LH (Chen, Long-Hui); Chen, YL (Chen, Yueh-Li); Chang, FMM (Chang, Fengming M.)
Title: Evaluating the process of a genetic algorithm to improve the back-propagation network:
A Monte Carlo study
Source: EXPERT SYSTEMS WITH APPLICATIONS, 36 (2): 1459-1465 Part 1 MAR 2009 Language: English
Document Type: Article
Author Keywords: Back-propagation network; Topology; Overfitting
KeyWords Plus: ARTIFICIAL NEURAL-NETWORKS; OPTIMIZATION; DESIGN; MODEL Abstract: Many studies have mapped a bit-string genotype using a genetic algorithm to represent network architectures to improve performance of back-propagation networks (BPN).
But the limitations of gradient search techniques applied to complex nonlinear optimization problems have often resulted in inconsistent and unpredictable performance. This study focuses oil how to collect and re-evaluate the weight matrices of a BPN while the genetic algorithm operations are processing in each generation to optimize the weight matrices. In this wily, overfitting. a drawback of BPNs that usually occurs during the later stage of neural network training with descending training error and ascending prediction error, can also be avoided. This study extends the parameters and topology of the neural network to enhance the feasibility of the solution space for complex nonlinear problems, The value of the proposed model is compared with previous studies Using a Monte Carlo study oil in-sample,
interpolation, and extrapolation data for six test functions. (c) 2007 Elsevier Ltd. All rights reserved.
Addresses: [Huang, Chien-Yu] Shu Te Univ, Dept Informat Management, Kaohsiung, Taiwan;
[Chen, Long-Hui] Shu Te Univ, Dept Logist Management, Kaohsiung, Taiwan; [Chen, Yueh-Li]
Cheng Shiu Univ, Dept Ind Engn & Management, Kaohsiung, Taiwan; [Chang, Fengming M.]
Asia Univ, Dept Informat Sci & Applicat, Taichung, Taiwan
Reprint Address: Huang, CY, Shu Te Univ, Dept Informat Management, Kaohsiung, Taiwan.
E-mail Address: [email protected] Funding Acknowledgement:
Funding Agency Grant Number
National Science Council, Taiwan, ROC NSC 95-2416-H-366-009
The authors gratefully acknowledge the financial support provided by the National Science Council, Taiwan, ROC under Grant No. NSC 95-2416-H-366-009.
Cited References: ARENA P, 1992, P 35 MIDW S CIRC SYS, V2, P1381.
CHANG PC, 2006, EXPERT SYST APPL, V30, P715, DOI 10.1016/j.eswa.2005.07.031.
GOLDBERG DE, 1989, GENETIC ALGORITHMS S.
GRACIA JD, 1997, ANAL CHIM ACTA, V348, P143.
GUPTA JND, 1999, OMEGA-INT J MANAGE S, V27, P679.
HAN J, 1996, ENG APPL ARTIF INTEL, V9, P109.
HECKERLING PS, 2004, ARTIF INTELL MED, V30, P71, DOI 10.1016/S0933- 3657(03)00065-4.
JIANG N, 2003, ADV ENG SOFTW, V34, P17.
KAMRANI AK, 2003, J INTELL MANUF, V14, P599.
KHAW JFC, 1995, NEUROCOMPUTING, V7, P225.
KIM KJ, 2000, EXPERT SYST APPL, V19, P125.
LI TS, 2003, COMPUT IND, V50, P113.
LIU YG, 2001, OPT COMMUN, V198, P369.
MACKEY MC, 1977, SCIENCE, V197, P287.
MANIEZZO V, 1994, IEEE T NEURAL NETWOR, V5, P39.
RIPLEY BD, 1993, CHAOS NETWORKS STAT.
SAMANTA B, 2003, ENG APPL ARTIF INTEL, V16, P657, DOI 10.1016/j.engappai.2003.09.006.
SCHITTENKOPF C, 1997, NEURAL NETWORKS, V10, P505.
SCHUMACHER M, 1996, COMPUT STAT DATA AN, V21, P661.
SEXTON RS, 1998, DECIS SUPPORT SYST, V22, P171.
SRINIVAS M, 1994, IEEE T SYST MAN CYB, V24, P656.
YASUMASA I, 1994, IEICE T FUND ELECTR, V77, P731.
ZHANG GP, 2001, COMPUT OPER RES, V28, P381.
ZHANG L, 1997, ANAL CHIM ACTA, V344, P29.
ZOU HF, 2007, EXPERT SYST APPL, V33, P347, DOI 10.1016/j.eswa.2006.05.021.
ZURADA JM, 1995, INTRO ARTIFICAL NEUR.
Cited Reference Count: 26 Times Cited: 1
Publisher: PERGAMON-ELSEVIER SCIENCE LTD
Publisher Address: THE BOULEVARD, LANGFORD LANE, KIDLINGTON, OXFORD OX5 1GB, ENGLAND
ISSN: 0957-4174
DOI: 10.1016/j.eswa.2007.11.055
29-char Source Abbrev.: EXPERT SYST APPL ISO Source Abbrev.: Expert Syst. Appl.
Source Item Page Count: 7
Subject Category: Computer Science, Artificial Intelligence; Engineering, Electrical &
Electronic; Operations Research & Management Science ISI Document Delivery No.: 390QE
