Comparison of Unsupervised Learning Algorithms for Clustering Cuban Citizens using a Lifestyle Questionnaire

Abstract

This study uses information technologies to analyze the lifestyles of Cubans. Cluster analysis is used to identify similarities in habits and lifestyles. Clustering results are compared using K-Means, DBSCAN, and HDBSCAN algorithms. Principal Component Analysis is applied to visualize the dataset. Internal validation metrics are defined to evaluate the performance of the algorithms. The results indicate that K-Means provides better clustering for this dataset.