Clustering algorithms subjected to K-mean and gaussian mixture model on multidimensional data set

Abstract

This paper explored the method of clustering. Two main categories of algorithms will be used, namely k-means and Gaussian Mixture Model clustering. We will look at algorithms within thesis categories and what types of problems they solve, as well as what methods could be used to determine the number of clusters. Finally, we will test the algorithms out using sparse multidimensional data acquired from the usage of a video games sales all around the world, we categories the sales in three main standards of high sales, medium sales and low sales, showing that a simple implementation can achieve nontrivial results. The result will be presented in the form of an evaluation of there is potential for online clustering of video games sales. We will also discuss some task specific improvements and which approach is most suitable.