出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：Classification and Prediction is an important research area of data mining. Construction ofclassifier model for any decision system is an important job for many data mining applications.The objective of developing such a classifier is to classify unlabeled dataset into classes. Herewe have applied a discrete Particle Swarm Optimization (PSO) algorithm for selecting optimalclassification rule sets from huge number of rules possibly exist in a dataset. In the proposedDPSO algorithm, decision matrix approach was used for generation of initial possibleclassification rules from a dataset. Then the proposed algorithm discovers important orsignificant rules from all possible classification rules without sacrificing predictive accuracy.The proposed algorithm deals with discrete valued data, and its initial population of candidatesolutions contains particles of different sizes. The experiment has been done on the task ofoptimal rule selection in the data sets collected from UCI repository. Experimental results showthat the proposed algorithm can automatically evolve on average the small number ofconditions per rule and a few rules per rule set, and achieved better classification performanceof predictive accuracy for few classes.
关键词：Particle swarm optimization; Data Mining; Classifiers.
