A data-splitting approach for comparing hierarchical clustering algorithms

• Grant number: 20/10861-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: November 01, 2020
• End date: October 31, 2021
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Statistics
• Principal Investigator: Rafael Izbicki
• Grantee: Luben Miguel Cruz Cabezas
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

Clustering methods have the goal of grouping together sample points that have similar properties. Several clustering techniques have been developed. An important question in practice is how to choose which method to use. While several approaches to answer this question have been proposed, they do not focus on hierarchical clustering methods because these do not yield a single partition of the observations. Thus, these evaluation techniques do not make use of the richer structure provided by these methods. In this project, we propose an alternative framework to compare clustering algorithms that is especially tailored for hierarchical methods. Our approach is also based on evaluating the predictive power of a given algorithm by using tools from phylogenetic analysis.

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

CABEZAS, LUBEN M. C.; IZBICKI, RAFAEL; STERN, RAFAEL B.. Hierarchical clustering: Visualization, feature importance and model selection. APPLIED SOFT COMPUTING, v. 141, p. 12-pg., 2023-07-01. (20/10861-7, 13/07699-0, 19/11321-9)