Choice of trimming proportion and number of clusters in robust clustering based on trimming

TL;DR

This paper provides a theoretical foundation and an automated bootstrap method for selecting the number of clusters and trimming proportion in robust clustering, reducing subjectivity in parameter choice.

Contribution

It introduces a theoretical understanding of classification trimmed likelihood curves and proposes a bootstrap approach to automate parameter selection.

Findings

Theoretical analysis of likelihood curves

Bootstrap method for parameter choice

Reduced set of sensible parameter options

Abstract

So-called "classification trimmed likelihood curves" have been proposed as a useful heuristic tool to determine the number of clusters and trimming proportion in trimming-based robust clustering methods. However, these curves needs a careful visual inspection, and this way of choosing parameters requires subjective decisions. This work is intended to provide theoretical background for the understanding of these curves and the elements involved in their derivation. Moreover, a parametric bootstrap approach is presented in order to automatize the choice of parameter more by providing a reduced list of "sensible" choices for the parameters. The user can then pick a solution that fits their aims from that reduced list.