Segmentation as a Tool of Data Analytics: A Methodological Comparison of Clustering Techniques for Quality of Life Indicators in European Countries

Ilias Grollios

Diplomová práce

Abstract:

This thesis compared clustering methodologies to categorize European countries based on data related to 15 Quality of Life indicators for the year 2022. These data were accessed from the EUROSTAT open database. The methodological comparison used Hierarchical Clustering with Complete Linkage and Ward's Method, as well as K-Means Clustering, in all cases with and without PCA, to effectively categorize countries. The use of PCA in clustering reduced dimensionality and improved clustering evaluation, as confirmed by the evaluation metrics of Silhouette Score, Davies-Bouldin Index, and Dunn Index. Hierarchical clustering employing the Ward Method in conjunction with PCA generated the most distinct and compact clusters. The study identified four clusters that revealed variations in European economic stability, employment, social engagement, and institutional trust. Emerging economies faced workforce challenges, whereas high-income countries possessed robust labour markets and educational institutions. Türkiye’s labour market inefficiencies and inadequate societal trust distinguished it from southern European countries with elevated unemployment rates. This data analytics study demonstrated how clustering may enhance understanding of regional inequities and inform policy initiatives.

Keywords

Quality of Life indicators, Clustering Evaluation, European Countries

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 3. 4. 2025

Obhajoba závěrečné práce

• Obhajoba proběhla 3. 6. 2025
• Vedoucí: doc. Ing. Tomáš Hlavsa, Ph.D.
• Oponent: Michal Kadlec, externi

Česká zemědělská univerzita v Praze

Provozně ekonomická fakulta

Magisterský studijní program:
Informatics