Content of Learning Activity

This course deals with the field at the intersection of artificial intelligence and probability theory: probabilistic graphical models, Bayesian networks, and their implementation through probabilistic programming. Decision-making in the face of uncertainty (missing information, noisy data, etc.) by statistical inference constitutes one of the major contributions of applied statistics to AI. This course will offer the following content:
- Theoretical reminders on the theory of probabilities and applied statistics,
- General theory of probabilistic graphical models and Bayesian networks,
- Theory of commonly used models: HMMs (hidden Markov models), particle filters, unsupervised grouping methods (clustering with continuous and discrete variables),
- Learning / inference methods with these models: inference and learning with so-called hidden variables (not measurable but important for classification or decision-making), Monte-Carlo simulation, Expectation-Maximization (EM) algorithm, IA and generative models.

This course will include practicals to acquire the theory.
 

Required Learning Resources/Tools

All learning resources and tools required for this cours are available via Moodle, the online e-learning platform of UMONS.

Recommended Learning Resources/Tools

Additional recommended material is also accessible through Moodle, the online e-learning platform of UMONS.

Other Recommended Reading

Not applicable

Mode of delivery

• Face-to-face

Type of Teaching Activity/Activities

• Cours magistraux
• Travaux pratiques
• Travaux de laboratoire
• Projet sur ordinateur

Evaluations

The assessment methods of the Learning Activity (AA) are specified in the course description of the corresponding Educational Component (UE)