Abstract:In thispaper, we willconsider an function. Withthetradition methods, thenumber objective function called theWeightedSum ofclusters maybepre-defined oritmaybesetby Validity Function (WSVF),whichisaweighted themethod.In thispaper,clustering by sum ofseveral normalized cluster validityminimization of criterion function willbe functions. In contrastto optimization considered. techniques intended tofinda single, global A related important issue totheclustering solution in a problemdomain,niching algorithms isthecluster validity criterion, which techniques havetheability tolocate multipledealswiththesignificance ofthestructure solutions inmultimodal domains. Hence,a imposed byaclustering algorithm. There aremany Niching BinaryParticle SwarmOptimizationcluster validity criteria available (1,(2, 8,9).The (NBPSO) approachis developedfor results oftheclustering algorithm depend onthe automatically constructing thepropernumber validity functions intensively, whichareusedto ofclusters aswellasappropriate partitioning of evaluate theindividual clusters. Since there isno thedataset.We alsohybridize theNBPSO perfect validity function that islikely toprovide methodwiththek-means algorithm tooptimizeconsistently reliable results across structures of theWSVF automatically. Inexperiments, we different problem instances, andwitha single showtheeffectiveness oftheWSVF andthe validity function, thealgorithm maybesusceptible validity oftheNBPSO.Incomparison with tolocal optima. Inthis paper, we address this otherrelated PSO,theNBPSOcanconsistently dilemma bysuggesting afunction called Weighted and efficiently convergeto theoptimum Sum Validity Function (WSVF),whichisa corresponding tothegiven datainconcurrenceweighted sumoftheseveral normalized validity withtheconvergence result. TheWSVF is functions, toconduct somesortofvoting among foundgenerally abletoimprove theconfidenceits component functions about thebest solution. of clustering solutions and achievemore Niching algorithms havetheability tolocate and accurate androbust results. maintain morethanonesolution toamulti-modal IndexTerms-Cluster validity, clustering, optimization problem. Whileniching hasbeen niching, particle swarmoptimization. studied extensively withreference togenetic algorithms (1), notmuchresearch hasbeendoneto

A Weighted Sum Validity Function for Clustering With BPSO Algorithm

Clustering with a Weighted Sum Validity Function Using a Niching PSO Algorithm

Clustering Withaweighted SumValidity Function Using Aniching PSOAlgorithm

A Dynamic Weighted Sum Validity Function for Fuzzy Clustering with an Adaptive Differential Evolution Algorithm

Particle Swarm Optimizer for Variable Weighting in Clustering High-Dimensional Data

A Fuzzy Similarity-Based Clustering Optimized by Particle Swarm Optimization

A Fuzzy Weighted Sum Validity Function for Clustering with a Mixed Strategy Evolutionary Algorithm

A New Particle Swarm Optimization Algorithm For Clustering

An Optimized K-means Clustering Algorithm Based on Improved Particle Swarm Optimization

Accelerated Two-Stage Particle Swarm Optimization for Clustering Not-Well-Separated Data

An adaptive particle swarm optimization method based on clustering

K-Means Clustering Algorithm Optimized By Particle Swarm Optimization Algorithm

Particle Swarm Optimizer For High-Dimensional Data Clustering

DYPSOKM: A Dynamic Union of PSO and K-Means, A Better Cluster

A Robust Adaptive Particle Swarm Optimization For Clustering Analysis Based On Steepest Descent Method

Particle Swarm Optimization Based Clustering: A Comparison of Different Cluster Validity Indices

Fuzzy C-Means with Improved Particle Swarm Optimization

Clustering Multidimensional Data with PSO based Algorithm

K-Means Cluster Algorithm Based on Improved PSO

A Novel Hybrid PSO-K-Means Clustering Algorithm Using Gaussian Estimation of Distribution Method and Lévy Flight

A Hybrid PSO Based on Dynamic Clustering for Global Optimization