Managing Population Diversity Through the Use of Weighted Objectives and Modified Dominance: An Example from Data Mining

Alan P. Reynolds; Beatriz de la Iglesia

doi:10.1109/MCDM.2007.369423

Managing Population Diversity Through the Use of Weighted Objectives and Modified Dominance: An Example from Data Mining

Alan P. Reynolds, Beatriz de la Iglesia

Research output: Contribution to conference › Paper

1 Citation (Scopus)

Abstract

The most successful multi-objective metaheuristics, such as NSGA II and SPEA 2, usually apply a form of elitism in the search. However, there are multi-objective problems where this approach leads to a major loss of population diversity early in the search. In earlier work, the authors applied a multi-objective metaheuristic to the problem of rule induction for predictive classification, minimizing rule complexity and misclassification costs. While high quality results were obtained, this problem was found to suffer from such a loss of diversity. This paper describes the use of both linear combinations of objectives and modified dominance relations to control population diversity, producing higher quality results in shorter run times

Original language	English
Pages	99-106
Number of pages	8
DOIs	https://doi.org/10.1109/MCDM.2007.369423
Publication status	Published - Apr 2007
Event	2007 IEEE Symposium on Computational Intelligence in Multicriteria Decision-Making - Honolulu, United States Duration: 1 Apr 2007 → 5 Apr 2007

Conference

Conference	2007 IEEE Symposium on Computational Intelligence in Multicriteria Decision-Making
Abbreviated title	MCDM 2007
Country/Territory	United States
City	Honolulu
Period	1/04/07 → 5/04/07

Access to Document

10.1109/MCDM.2007.369423

Cite this

@conference{0ae3e45438eb48a7a442e5e79b803898,

title = "Managing Population Diversity Through the Use of Weighted Objectives and Modified Dominance: An Example from Data Mining",

abstract = "The most successful multi-objective metaheuristics, such as NSGA II and SPEA 2, usually apply a form of elitism in the search. However, there are multi-objective problems where this approach leads to a major loss of population diversity early in the search. In earlier work, the authors applied a multi-objective metaheuristic to the problem of rule induction for predictive classification, minimizing rule complexity and misclassification costs. While high quality results were obtained, this problem was found to suffer from such a loss of diversity. This paper describes the use of both linear combinations of objectives and modified dominance relations to control population diversity, producing higher quality results in shorter run times",

author = "Reynolds, {Alan P.} and {de la Iglesia}, Beatriz",

year = "2007",

month = apr,

doi = "10.1109/MCDM.2007.369423",

language = "English",

pages = "99--106",

note = "2007 IEEE Symposium on Computational Intelligence in Multicriteria Decision-Making, MCDM 2007 ; Conference date: 01-04-2007 Through 05-04-2007",

}

TY - CONF

T1 - Managing Population Diversity Through the Use of Weighted Objectives and Modified Dominance: An Example from Data Mining

AU - Reynolds, Alan P.

AU - de la Iglesia, Beatriz

PY - 2007/4

Y1 - 2007/4

N2 - The most successful multi-objective metaheuristics, such as NSGA II and SPEA 2, usually apply a form of elitism in the search. However, there are multi-objective problems where this approach leads to a major loss of population diversity early in the search. In earlier work, the authors applied a multi-objective metaheuristic to the problem of rule induction for predictive classification, minimizing rule complexity and misclassification costs. While high quality results were obtained, this problem was found to suffer from such a loss of diversity. This paper describes the use of both linear combinations of objectives and modified dominance relations to control population diversity, producing higher quality results in shorter run times

AB - The most successful multi-objective metaheuristics, such as NSGA II and SPEA 2, usually apply a form of elitism in the search. However, there are multi-objective problems where this approach leads to a major loss of population diversity early in the search. In earlier work, the authors applied a multi-objective metaheuristic to the problem of rule induction for predictive classification, minimizing rule complexity and misclassification costs. While high quality results were obtained, this problem was found to suffer from such a loss of diversity. This paper describes the use of both linear combinations of objectives and modified dominance relations to control population diversity, producing higher quality results in shorter run times

U2 - 10.1109/MCDM.2007.369423

DO - 10.1109/MCDM.2007.369423

M3 - Paper

SP - 99

EP - 106

T2 - 2007 IEEE Symposium on Computational Intelligence in Multicriteria Decision-Making

Y2 - 1 April 2007 through 5 April 2007

ER -