SIMULTANEOUS MODEL FOR CLUSTERING AND INTRA-GROUP FEATURE SELECTION

Abstract

In this thesis, we focus on developing an algorithmic framework to perform clustering and intra-group feature selection simultaneously. In order to achieve this goal, we first study the convex clustering model and the exclusive lasso model in Chapter 3 and Chapter 4, respectively. Then, we study the new sparse convex clustering model in Chapter 5, which can achieve the goal of performing clustering and data point wise feature selection simultaneously. In summary, this thesis contributes to the topic of clustering and intra-group level feature selection from both the model analysis and numerical optimization algorithm perspectives.