Open Access Open Access  Restricted Access Subscription or Fee Access

Cluster Detection through Model Based Clustering Approach

Vikas Singh, Mukesh Kumar

Abstract



Cluster analysis is an important technique for creating homogeneous group of objects, items. These groupings are sometimes essential for the proper analysis of data. Therefore, the technique of clustering is very useful in data analysis, especially in case of large data-sets. In clustering, the groups or clusters are found based on some characteristics of interest. Based on the chosen characteristics of interest similarity measures/distance matrices are calculated to perform the cluster analysis through heuristic clustering techniques. However, a more scientific approach of clustering is model based clustering approach. This approach is basically based on the distributional setups of observations. In this method, the data points are supposed to be drawn from a mixture distribution. The mixture distribution of multivariate normal gives some exciting and interesting result as compared to the other component distributions. In this paper, we have made an attempt to find clusters through a mixture of multivariate normal as components for the real dataset of student enrollment in higher education in India for clustering. The datasets is taken from All India Survey on Higher Education (AISHE) report for the year 2017-18. BIC is used for reporting the optimal no. of cluster and adequacy of the model.

Keywords


Model based cluster, EM-algorithm, Datasets, Bayesian Information Criterion.

Full Text:

PDF


Disclaimer/Regarding indexing issue:

We have provided the online access of all issues and papers to the indexing agencies (as given on journal web site). It’s depend on indexing agencies when, how and what manner they can index or not. Hence, we like to inform that on the basis of earlier indexing, we can’t predict the today or future indexing policy of third party (i.e. indexing agencies) as they have right to discontinue any journal at any time without prior information to the journal. So, please neither sends any question nor expects any answer from us on the behalf of third party i.e. indexing agencies.Hence, we will not issue any certificate or letter for indexing issue. Our role is just to provide the online access to them. So we do properly this and one can visit indexing agencies website to get the authentic information.