Data Mining Rules Using Multi-Objective Evolutionary Algorithms

B. de la Iglesia; M. S. Philpott; A. J. Bagnall; V. J. Rayward-Smith

doi:10.1109/CEC.2003.1299857

Data Mining Rules Using Multi-Objective Evolutionary Algorithms

B. de la Iglesia, M. S. Philpott, A. J. Bagnall, V. J. Rayward-Smith

Research output: Contribution to conference › Paper

53 Citations (Scopus)

Abstract

In data mining, nugget discovery is the discovery of interesting classification rules that apply to a target class. In previous research, heuristic methods (genetic algorithms, simulated annealing and tabu search) have been used to optimise a single measure of interest. This paper proposes the use of multi-objective optimisation evolutionary algorithms to allow the user to interactively select a number of interest measures and deliver the best nuggets (an approximation to the Pareto-optimal set) according to those measures. Initial experiments are conducted on a number of databases, using an implementation of the fast elitist non-dominated sorting genetic algorithm (NSGA), and two well known measures of interest. Comparisons with the results obtained using modern heuristic methods are presented. Results indicate the potential of multi-objective evolutionary algorithms for the task of nugget discovery.

Original language	English
Pages	1552-1559
Number of pages	8
DOIs	https://doi.org/10.1109/CEC.2003.1299857
Publication status	Published - Dec 2003
Event	2003 IEEE Congress on Evolutionary Computation - Canberra, Australia Duration: 8 Dec 2003 → 12 Dec 2003

Conference

Conference	2003 IEEE Congress on Evolutionary Computation
Country/Territory	Australia
City	Canberra
Period	8/12/03 → 12/12/03

Access to Document

10.1109/CEC.2003.1299857

Cite this

@conference{f3679587f0264e08b8c5d39bf5f7f431,

title = "Data Mining Rules Using Multi-Objective Evolutionary Algorithms",

abstract = "In data mining, nugget discovery is the discovery of interesting classification rules that apply to a target class. In previous research, heuristic methods (genetic algorithms, simulated annealing and tabu search) have been used to optimise a single measure of interest. This paper proposes the use of multi-objective optimisation evolutionary algorithms to allow the user to interactively select a number of interest measures and deliver the best nuggets (an approximation to the Pareto-optimal set) according to those measures. Initial experiments are conducted on a number of databases, using an implementation of the fast elitist non-dominated sorting genetic algorithm (NSGA), and two well known measures of interest. Comparisons with the results obtained using modern heuristic methods are presented. Results indicate the potential of multi-objective evolutionary algorithms for the task of nugget discovery.",

author = "{de la Iglesia}, B. and Philpott, {M. S.} and Bagnall, {A. J.} and Rayward-Smith, {V. J.}",

year = "2003",

month = dec,

doi = "10.1109/CEC.2003.1299857",

language = "English",

pages = "1552--1559",

note = "2003 IEEE Congress on Evolutionary Computation ; Conference date: 08-12-2003 Through 12-12-2003",

}

TY - CONF

T1 - Data Mining Rules Using Multi-Objective Evolutionary Algorithms

AU - de la Iglesia, B.

AU - Philpott, M. S.

AU - Bagnall, A. J.

AU - Rayward-Smith, V. J.

PY - 2003/12

Y1 - 2003/12

N2 - In data mining, nugget discovery is the discovery of interesting classification rules that apply to a target class. In previous research, heuristic methods (genetic algorithms, simulated annealing and tabu search) have been used to optimise a single measure of interest. This paper proposes the use of multi-objective optimisation evolutionary algorithms to allow the user to interactively select a number of interest measures and deliver the best nuggets (an approximation to the Pareto-optimal set) according to those measures. Initial experiments are conducted on a number of databases, using an implementation of the fast elitist non-dominated sorting genetic algorithm (NSGA), and two well known measures of interest. Comparisons with the results obtained using modern heuristic methods are presented. Results indicate the potential of multi-objective evolutionary algorithms for the task of nugget discovery.

AB - In data mining, nugget discovery is the discovery of interesting classification rules that apply to a target class. In previous research, heuristic methods (genetic algorithms, simulated annealing and tabu search) have been used to optimise a single measure of interest. This paper proposes the use of multi-objective optimisation evolutionary algorithms to allow the user to interactively select a number of interest measures and deliver the best nuggets (an approximation to the Pareto-optimal set) according to those measures. Initial experiments are conducted on a number of databases, using an implementation of the fast elitist non-dominated sorting genetic algorithm (NSGA), and two well known measures of interest. Comparisons with the results obtained using modern heuristic methods are presented. Results indicate the potential of multi-objective evolutionary algorithms for the task of nugget discovery.

U2 - 10.1109/CEC.2003.1299857

DO - 10.1109/CEC.2003.1299857

M3 - Paper

SP - 1552

EP - 1559

T2 - 2003 IEEE Congress on Evolutionary Computation

Y2 - 8 December 2003 through 12 December 2003

ER -