Room “Seminar Room” - Abacus Building (ex U14)
Â
Label ranking with Bayesian Networks and Decision Trees
Â
Speakers: Prof. José A. Gámez and José M. Puerta, University of Castilla-La Mancha (Spain)
Â
Abstract
Label ranking is a machine learning task in which the goal is to predict an ordered list of labels rather than a single class. This seminar presents several approaches to label ranking based on decision trees and Bayesian networks, with special attention to the case of partial rankings, where only incomplete preference information is available and ties are allowed. We will discuss how these models can represent and learn preference structures from data, and how they deal with uncertainty and dependencies among labels. The seminar will also summarize the main empirical results obtained in our work and highlight some future research.
Â
Short bios
José A. Gámez is a Full Professor at the Dpt. of Computing Systems at the University of Castilla-La Mancha (Albacete, Spain).  His research interests include probabilistic reasoning, Bayesian networks, evolutionary algorithms, machine learning, and data mining. He has authored and coauthored more than 80 papers in journals and more than 150 in international conferences. He has collaborated in the edition of several books, proceedings, and special issues of international journals. He currently co-leads the Research Group on Intelligent Systems and Data Mining at UCLM.
José M. Puerta is a Full Professor at the Dpt. of Computing Systems at the University of Castilla-La Mancha (Albacete, Spain). His research interests include probabilistic reasoning, probabilistic graphical models, Bayesian networks, machine learning, and explainable artificial intelligence. He has authored and coauthored more than 100 papers in international journals and conferences. Dr. Puerta has served as Guest Editor for special issues of international journals and co-leads the Research Group on Intelligent Systems and Data Mining at UCLM.
Contact person for the seminar: fabio.stella@unimib.it