Uncertain Reasoning in Al

3 credits

Advances in using uncertain knowledge to make decisions rationally and effectively (for prediction, classification, diagnosis, etc.). Focus is placed on modelling with Bayesian networks and various methods for inference therein. Deep learning probabilistic circuits, including arithmetic circuits and sum-product networks, will be extensively studied, emphasizing construction, learning, semantics, and inference. *Note: Before registering for this course, students should have a background in discrete computational structures, artificial intelligence and statistical methods comparable to the senior undergraduate level.*