Faculty Advisor

Neil T. Heffernan

Faculty Advisor

Gabor N. Sarkozy

Faculty Advisor

Sonia Chernova


The fact that clustering is perhaps the most used technique for exploratory data analysis is only a semaphore that underlines its fundamental importance. The general problem statement that broadly describes clustering as the identification and classification of patterns into coherent groups also implicitly indicates it's utility in other tasks such as supervised learning. In the past decade and a half there have been two developments that have altered the landscape of research in clustering: One is improved results by the increased use of graph theoretic techniques such as spectral clustering and the other is the study of clustering with respect to its relevance in semi-supervised learning i.e. using unlabeled data for improving prediction accuracies. In this work an attempt is made to make contributions to both these aspects. Thus our contributions are two-fold: First, we identify some general issues with the spectral clustering framework and while working towards a solution, we introduce a new algorithm which we call "Regularity Clustering" which makes an attempt to harness the power of the Szemeredi Regularity Lemma, a remarkable result from extremal graph theory for the task of clustering. Secondly, we investigate some practical and useful strategies for using clustering unlabeled data in boosting prediction accuracy. For all of these contributions we evaluate our methods against existing ones and also apply these ideas in a number of settings.


Worcester Polytechnic Institute

Degree Name



Computer Science

Project Type


Date Accepted





Machine Learning, Graph Mining, Unsupervised Learning, Ensemble Learning, Semi-Supervised Learning, Regularity Lemma, Graph Partitioning