Hierarchical Gaussian Mixture Model with Objects Attached to Terminal and Non-terminal Dendrogram Nodes

TL;DR

This paper introduces a hierarchical Gaussian mixture model that uniquely stores objects in both terminal and non-terminal nodes, improving noise detection and producing more compact, higher-quality dendrograms.

Contribution

It presents a novel hierarchical clustering method that enhances object storage flexibility and dendrogram quality compared to traditional models.

Findings

Improved noise detection through hierarchical structure.

More compact dendrograms with higher F-measure.

Effective modeling of objects at different hierarchy levels.

Abstract

A hierarchical clustering algorithm based on Gaussian mixture model is presented. The key difference to regular hierarchical mixture models is the ability to store objects in both terminal and nonterminal nodes. Upper levels of the hierarchy contain sparsely distributed objects, while lower levels contain densely represented ones. As it was shown by experiments, this ability helps in noise detection (modelling). Furthermore, compared to regular hierarchical mixture model, the presented method generates more compact dendrograms with higher quality measured by adopted F-measure.