Neural Compression — Lecture 4 — Random Variables and Autoregressive Models

Fourth week of the course “Data Compression With Deep Probabilistic Models“ by Robert Bamler at University of Tübingen (summer term of 2021). Viewers who are already familiar with probability theory may consider skipping the interlude from 44:00 to 1:51:00. Playlist for entire course: Lecture notes: Problem Set 3 (implement a compression method with a learned autoregressive model): Problem Set 4 (conditional entropies): Solutions to Problem Set 3: Solutions to Problem Set 4: More course materials: © Robert Bamler / University of Tübingen, 2021 C