Rough set approximate entropy reducts with order based particle swarm optimization

Authors:
Xiangyang Wang;Wanggen Wan;Xiaoqing Yu
Affiliations:
Shanghai University, Shanghai, China, Shanghai, China;Shanghai University, Shanghai, China, Shanghai, China;Shanghai University, Shanghai, China, Shanghai, China
Venue:
Proceedings of the first ACM/SIGEVO Summit on Genetic and Evolutionary Computation
Year:
2009

Citing 9
Cited 0

Knowledge discovery in databases: an attribute-oriented rough set approach

Knowledge discovery in databases: an attribute-oriented rough set approach
Theoretical foundations of order-based genetic algorithms

Fundamenta Informaticae - Special issue: to the memory of Prof. Helena Rasiowa
Rough set algorithms in classification problem

Rough set methods and applications
Rough Sets: Theoretical Aspects of Reasoning about Data

Rough Sets: Theoretical Aspects of Reasoning about Data
Rough set methods in feature selection and recognition

Pattern Recognition Letters - Special issue: Rough sets, pattern recognition and data mining
Theoretical Study on Attribute Reduction of Rough Set Theory: Comparison of Algebra and Information Views

ICCI '04 Proceedings of the Third IEEE International Conference on Cognitive Informatics
Ensembles of Classifiers Based on Approximate Reducts

Fundamenta Informaticae - Concurrency Specification and Programming (CS&P'2000)
Feature selection based on rough sets and particle swarm optimization

Pattern Recognition Letters
Order based genetic algorithms for the search of approximate entropy reducts

RSFDGrC'03 Proceedings of the 9th international conference on Rough sets, fuzzy sets, data mining, and granular computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose an order-based Particle Swarm Optimization (o-PSO) hybrid algorithm for rough set approximate entropy reducts (oPSOAER). The o-PSO generates proper permutation of attributes, which are used by approximate entropy reduction algorithm to produce rough set reducts. The reducts are evaluated by fitness function. The primary criterion of optimization of the fitness function is the number of rules and the secondary is the reduct length. Our algorithm is tested on some UCI datasets. The results show that oSPOAER is efficient for approximate entropy reducts. The approximate entropy reducts optimized according to number of rules are better in classification algorithms than the shortest ones, and are much better for practical applications.