Supervised and Unsupervised Statistical Data Analysis

This book contains peer-reviewed contributions from the 15th Scientific Meeting of the Classification and Data Analysis Group (CLADAG) of the Italian Statistical Society (SIS), held in Naples, Italy, from 8 to 10 September 2025. This edition marks a historic milestone: for the first time, CLADAG has hosted a full joint scientific meeting with another classification society—the Vereniging voor Ordinatie en Classificatie (VOC), the Dutch-Flemish classification society. Over the years, CLADAG has organized other joint scientific events with classifi-cation societies, such as the Société Francophone de Classification (SFC) in 2008, the German Classification Society (GfKl) in 2010, and the Japanese Classification Society (JCS) in 2012. However, those meetings were not part of the official CLADAG biennial cycle and were instead smaller, intermediate events. The 2025 edition stands out as the first time CLADAG and VOC have co-organized an official conference of both societies. The CLADAG-VOC 2025 conference brings together two long-standing traditions in statistical science. Since its founding, CLADAG has been Italy’s leading platform for advancing methodological research in multivariate statistics, particularly in classification and data analysis. A member of the International Federation of Classification Societies (IFCS) since 1985, CLADAG has promoted scientific progress through biennial confer-ences hosted across Italy—from Pescara in 1997 to Salerno in 2023—as well as through specialized schools and collaborative initiatives with other IFCS members. This joint meeting takes on special significance through the collaboration with VOC, which has cultivated an equally distinguished tradition since its establishment in 1989. Also, a founding member of the IFCS, VOC has consistently supported innovative research in data analysis, classification, clustering methodology, and ordination, both through its annual meetings and international collaborations. The intellectual synergy created by the joint CLADAG-VOC meeting elevates this volume beyond standard con-ference proceedings. It exemplifies how two societies can connect theoretical innovation with practical application. The selected papers demonstrate how international collabora-tion can generate novel solutions to long-standing challenges in statistical data analysis and address the emerging needs of an increasingly complex data landscape. The Sci-entific Committee of CLADAG-VOC 2025 designed the keynote sessions to provide fresh perspectives on the current state of research in clustering, classification, and data analysis. The scientific program is particularly rich, featuring five keynote lectures, 42 invited sessions sponsored by members of the scientific committee, and, at the time of writing, 14 contributed sessions. We wish to express our deepest gratitude to all participants—authors, reviewers, session chairs, speakers, and committee members—whose dedication made this joint meeting possible. Special thanks goes to the Italian Statistical Society (SIS) and the Dutch-Flemish Classification Society (VOC) for their institutional support, to the Uni-versity of Naples Federico II and its Department of Economics and Statistics for both institutional and financial backing, and to Springer for its continued collaboration in dis-seminating research. We hope that this volume will serve as a testament to the value of international collaboration and inspire future breakthroughs at the forefront of statistical science. This volume captures the intellectual vitality and collaborative spirit that charac-terized the meeting, showcasing how the convergence of distinct statistical traditions can enrich data analysis. The contributions offer new insights into the theoretical and prac-tical challenges of supervised and unsupervised learning, highlighting the remarkable breadth of contemporary statistical research while maintaining methodological rigour. Innovative approaches to statistical modelling—addressing spatial dependencies and circular data structures—are presented alongside significant advances in interpretable machine learning that reconcile statistical precision with algorithmic transparency. Par-ticularly noteworthy is the volume’s treatment of complex data structures, including novel methods for network analysis, high-dimensional clustering, temporal pattern recognition, optimization techniques. What distinguishes this volume is its seamless integration of methodological innovation and practical relevance. The applications span diverse domains, including the social sciences and biomedical engineering, each demon-strating the effective translation of statistical theory into real-world impact. The 30 selected papers collectively push the boundaries of the field: reconciling traditional sta-tistical approaches with modern computational demands, developing new tools for inter-preting increasingly complex data, and illustrating how well-crafted methodologies can address substantive questions across multiple disciplines. This interdisciplinary char-acter reflects the unique value of international collaborations like the CLADAG-VOC partnership, and we are truly grateful for it. As editors, we hope this volume will serve both as a snapshot of current innovations at the intersection of classification and data analysis, and as an inspiration for future research and collaboration. June 2025 Antonio D’Ambrosio Mark de Rooij Kim De Roover Carmela Iorio Michele La Rocca