CRAFT: ClusteR-specific Assorted Feature selecTion

TL;DR

CRAFT is a clustering framework that performs cluster-specific feature selection, effectively handling mixed data types, and offers a scalable, easy-to-implement solution that does not require predefining the number of clusters.

Contribution

It introduces a novel nonparametric MAP-based clustering approach with cluster-specific feature selection for mixed data types, improving flexibility and ease of use.

Findings

Performs well on real datasets

Scales efficiently to large data

Requires minimal parameter tuning

Abstract

We present a framework for clustering with cluster-specific feature selection. The framework, CRAFT, is derived from asymptotic log posterior formulations of nonparametric MAP-based clustering models. CRAFT handles assorted data, i.e., both numeric and categorical data, and the underlying objective functions are intuitively appealing. The resulting algorithm is simple to implement and scales nicely, requires minimal parameter tuning, obviates the need to specify the number of clusters a priori, and compares favorably with other methods on real datasets.