WSEAS Transactions on Computers
Print ISSN: 1109-2750, E-ISSN: 2224-2872
Volume 17, 2018
FGPSO - A Novel Algorithm for Multi Objective Data Clustering
Authors: ,
Abstract: The task of clustering is to group the data items that are similar into different clusters in such a way that the similarity within each cluster is high and the dissimilarity between the clusters is also high. A novel partitional clustering algorithm called HB K-Means algorithm (High Dimensional Bisecting K-Means) based on high dimensional data set was developed in our previous work. In order to improve this novel algorithm, constraints such as Stability based measure and Mean Square Error (MSE) were incorporated resulting in CHB K-Means (Constraint Based HB K-Means) algorithm. In addition to these constraints, cluster compactness and density are also important to obtain better clustering results. In this paper, it is proposed to develop a Multi-Objective Optimization (MOO) technique by including different indices such as DB-Index, XB-Index and Sym-Index. These three indices will be used as fitness function for the proposed Fractional Genetic PSO algorithm (FGPSO) which is the hybrid optimization algorithm to do the clustering process. The performance of this optimization algorithm is evaluated based on parameters such as Clustering Accuracy and Time Computation by executing the algorithm on some of the benchmark datasets taken from UCI Machine Learning Repository.
Search Articles
Keywords: Partitional Clustering, Multi-Objective Optimization, DB Index, XB-Index, Sym-index, Fractional Genetic PSO Algorithm (FGPSO)
Pages: 1-9
WSEAS Transactions on Computers, ISSN / E-ISSN: 1109-2750 / 2224-2872, Volume 17, 2018, Art. #1