Séminaire (organisé par l’équipe de recherche ASER)

Alessandro ANTONUCCI

Chercheur au "Dalle Molle Institute for Artificial Intelligence (IDSIA)", Lugano - Suisse

Bayesian networks with imprecise probabilities : theory and applications to knowledge-based systems and classification

Mardi 10 décembre à 14h00 en salle C221

Résumé :

Bayesian networks are important tools for uncertain reasoning in aI ; their quantification requires a precise assessment of the conditional probabilities. Credal networks generalize Bayesian networks, so that probabilities can vary in a set (e.g., interval). This provides a more realistic model of expert knowledge and returns more robust inferences. We first outline the specification procedure of a credal network and the main approaches to decision making. A prototypical example of knowledge-based expert system related to military decisions based on credal networks is indeed presented. We also describe the major examples of credal classifiers, i.e., classification algorithms based on credal networks, developed so far. Credal classifiers allow robust classification even with missing data and can return more classes if the assignment to a single class is too uncertain ; in this way, they preserve reliability.


FR SHIC 3272

Collegium UTC/CNRS