We have built a new Modified Particle Swarm Optimization algorithm and worked on multi-dimensional clustering. The aim of the project is to get rid of the problems of standard PSO and firster convergence velocity. Clustering is a popular data analysis technique to identify homogeneous groups of objects based on the values of their attributes. Ten typical well known multi-dimensional data sets from the UCI machine learning repository are used to demonstrate the results of the techniques. We have tested these datasets using five fitness functions as DB Index, SI Index, Quantization Error, Sum of Square Error, CH Index. Modified Particle Swarm Optimization algorithm is efficient enough in global search and also efficient enough to handle the accurate clustering result. In our algorithm, the problems of the standard PSO algorithm are solved. Our algorithm gives better performance in all aspect.
