東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Neural enhancement for multiobjectiv...

Auburn University.

Linked to FindBook

Google Book

Amazon

博客來

Neural enhancement for multiobjective optimization.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Neural enhancement for multiobjective optimization./
Author:	Garrett, Aaron.
Description:	219 p.
Notes:	Adviser: Gerry Dozier.
Contained By:	Dissertation Abstracts International69-06B.
Subject:	Computer Science. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3317313
ISBN:	9780549668107

Neural enhancement for multiobjective optimization.
Garrett, Aaron.

Neural enhancement for multiobjective optimization. - 219 p.

Adviser: Gerry Dozier.

Thesis (Ph.D.)--Auburn University, 2008.

In this work, a neural network approach is applied to multiobjective optimization problems in order to expand the set of optimal solutions. The network is trained using results obtained from existing evolutionary multiobjective optimization approaches. The network is then evaluated based on its performance against those same approaches when given more processing time. The results are collected from a set of well-known benchmark multiobjective problems, and its performance is evaluated using various indicators from the multiobjective optimization literature.

ISBN: 9780549668107Subjects--Topical Terms:

626642
Computer Science.

Neural enhancement for multiobjective optimization.
LDR:01952nmm 2200277 a 45 001 890856
005 20101105
008 101105s2008 ||||||||||||||||| ||eng d
020 $a 9780549668107
035 $a (UMI)AAI3317313
035 $a AAI3317313
040 $a UMI $c UMI
100 1 $a Garrett, Aaron. $3 724236
245 1 0 $a Neural enhancement for multiobjective optimization.
300 $a 219 p.
500 $a Adviser: Gerry Dozier.
500 $a Source: Dissertation Abstracts International, Volume: 69-06, Section: B, page: 3654.
502 $a Thesis (Ph.D.)--Auburn University, 2008.
520 $a In this work, a neural network approach is applied to multiobjective optimization problems in order to expand the set of optimal solutions. The network is trained using results obtained from existing evolutionary multiobjective optimization approaches. The network is then evaluated based on its performance against those same approaches when given more processing time. The results are collected from a set of well-known benchmark multiobjective problems, and its performance is evaluated using various indicators from the multiobjective optimization literature.
520 $a Preliminary experiments reveal the viability of this approach for expanding the set of solutions to multiobjective problems. Further experiments prove that it is possible to train the neural network in a reasonable time using heuristic methods. The results of this training approach are shown to be very competitive with the underlying evolutionary multiobjective optimization approach that was used to produce the training set. Additional experiments reveal the applicability of this approach across existing multiobjective optimization approaches.
590 $a School code: 0012.
650 4 $a Computer Science. $3 626642
690 $a 0984
710 2 $a Auburn University. $3 1020457
773 0 $t Dissertation Abstracts International $g 69-06B.
790 $a 0012
790 1 0 $a Dozier, Gerry, $e advisor
791 $a Ph.D.
792 $a 2008
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3317313