Vibrating Particle System Algorithm for Healthcare Datasets

Abstract

Clustering is a tool in data mining which used to obtain the information which is hidden from large sets of data of various structures and clusters. It is has been seen that in the domain of engineering that over the years the focus has now shifted to the use of computing techniques which draw their inspiration from nature. The report puts forward a latest meta heuristic technique, i.e. the Vibrating-Particle System (i.e. VPS) to solve various issues of global optimization. It is essentially a meta heuristic algorithm based on population as well as the damped free vibration of single-degree of freedom system. On a set of five standard datasets which are pertaining to medical care from the UCI-Machine Learning-Repository, we evaluated the proposed algorithm. The overall results which are obtained after implementation and computations have indicated that in terms of calculations and accuracy measures, Vibrating-Particle System outshines the other hi-tech and advanced algorithms.