Congratulations to CS Ph.D. Student Tal Friedman: his paper on “Approximate Knowledge Compilation by Online Collapsed Importance Sampling” was selected for full oral presentation at the Thirty-second Conference on Neural Information Processing Systems (NeurIPS). Tal’s paper was one of only 30 papers selected from 4856 submissions to give a talk at the main conference.
NeurIPS is the leading conference in machine learning and takes place next week in Montreal, Canada. Tickets to attend the conference sold out in less than 12 minutes.
Tal is a member of the Statistical and Relational Artificial Intelligence (StarAI) lab, advised by Prof. Guy Van den Broeck. His NeurIPS paper proposes a novel approximate inference algorithm for discrete probabilistic graphical models. Tal’s research is on techniques and models for probabilistic reasoning, at the intersection of probability and logic. In particular, he works on probabilistic databases, probabilistic programming, graphical models, and knowledge compilation.