Artificial Intelligence Group
The Artificial Intelligence Group of the Computer Science Institute of U.P.O. is the biggest group of the institute. Several AI research topics are investigated by the researchers involved in the group's activity: machine learning, data mining, knowledge represention and inference, case-based reasoning.
Machine Learning and Data Mining
The machine learning and data mining group is active in the area of Supervised and Unsupervised Learning, Deep Learning and in learning Probabilistic Graphical Models.
Case-Based Reasoning (CBR) involves reasoning from past cases and episodes to solve new problems. The CBR group currently investigates different aspects of the paradigm: adaptation, distance-based and fuzzy retrieval (with particular emphasis on Fuzzy-SQL retrieval), case base maintenance. Applications issues are also investigated with particular attention to (diabetes and dialysis) and e-commerce.
The Temporal Reasoning group is working on several topics concerning the management of temporal information in data and knowledge bases. Languages for user-defined periodicity: uniform analysis of existing approaches, driven from logical interpretations; semantical analysis of periodicities through appropriate representations; study and classifications of periodicities, and design of minimal languages. Temporal dimensions in Model-Based Reasoning and Diagnosis, especially reasoning on dynamic system, modeling and reasoning on events and change.
Probabilistic Reasoning and Bayesian Networks
The major focus of the group is to investigate the capabilities offered by probabilistic models (and in particular Bayesian Networks) for uncertain reasoning applications. Particular attention is devoted to the application of Bayesian Networks in Reliability application (see also the corresponding page ) and in diagnostic reasoning. Learning aspects of Bayesian Networks Classifiers are another area of interest.
Logic for Artificial Intelligence and Automated Reasoning
Reasoning about actions and change. Belief revision. Multi-agent systems. Deduction methods for modal and conditional logics.
- PRIN '03: Sviluppo e verifica di sistemi multiagente basati sulla logica
- Networks of Excellence: REWERSE, CoLogNet
Qualitative modeling and reasoning for simulation and diagnosis. Use of constraint reasoning and constraint languages for Model-Based Reasoning. Industrial applications of Model-Based Diagnosis, and MBD in embedded systems.