Classification and Data Analysis Group
Società Italiana di Statistica

Statistical Methods for Unsupervised and Supervised Learning with Dimensionality Reduction
- Stimulate collaborations and develop research together -

Rome (Italy), February 19-23, 2024

In today's information age, huge amounts of data, often with a complex multidimensional structure, are available in many fields. To learn from such data, statistical methods of classification, clustering and dimensionality reduction play a key role.
The international CLADAG (CLAssification and Data Analysis Group of the Italian Statistical Society) school aims to present, analyse and discuss the statistical methods for unsupervised and supervised learning with dimensionality reduction. The course is especially dedicated, but not limited to Master's students, doctoral students and postdoctoral researchers.
Starting from basic concepts, the course will focus on novel techniques and software for unsupervised and supervised learning with dimensionality reduction through extensive applications to real data.
During the lectures and especially on the last day, the course will be dedicated to the ongoing research activity of the lecturers and, possibly, of the participants on the topics of the course to stimulate possible collaborations.

Lecturers
Michael Greenacre, Pompeu Fabra University; Agostino Di Ciaccio, Sapienza University of Rome;
Paolo Giordani, Sapienza University of Rome; Roberto Rocci, Sapienza University of Rome; Maurizio Vichi, Sapienza University of Rome

Topics
Dimensionality reduction for quantitative data; Dimensionality reduction for categorical data; Fuzzy unsupervised classification; Model-based (unsupervised) classification; Supervised Classification

Software
Matlab; Python; R

The 14th Scientific Meeting of the Classification and Data Analysis Group (CLADAG) of the Italian Statistical Society (SIS) was held on 11 - 13 September 2023 at the "Fisciano Campus" in Salerno (CLADAG2023). A vibrant climate of scientific discussion has been built among around 300 participants, with 5 plenary talks, 31 invited sessions, and 27 spontaneous sessions. It was an enrichment of research perspectives for each of us who were present. Also exceptional was the participation of 120 foreign colleagues, who continue to contribute to the prestige of the Cladag conferences. Further, the Portuguese (CLAD), French (SFC), and German (GFKL) classification societies and the International Biometric Society (Italian region) actively contributed to the organization of the workshop.

The Scientific Committee of Cladag 2023, chaired by Carla Rampichini, conveyed remarkable scientific contributions. The local organizing committee, composed of Pietro Coretto, Giuseppe Giordano, Marcella Niglio, Maria Lucia Parrella, Marialuisa Restaino, Paolo Mascambruno, Domenico Vistocco and Maria Prosperina Vitale, was chaired by Michele la Rocca: thank you to all of them for making the event possible and for letting us enjoy it.

Keynote lecturers were given by:

• Francesco Bartolucci, University of Perugia, Italy, on discrete latent variable models
• Gerda Claeskens, Katholieke Universiteit Leuven, Belgium, on randomized group lasso
• Francesca Greselin, University of Milano-Bicocca, Italy, on robust clustering methods
• Thomas Kneib, Georg-August-Universität Göttingen, Germany, on distributional regression
• Sofia Charlotta Olhede, École Polytechnique Fédérale de Lausanne, Switzerland, on models for interaction data.

The next Scientific Meeting "CLADAG2025" will be held in Sardinia (Italy).