Advances in Distribution Compression by Lester Mackey
Date: December 7, 2022
Speaker: Lester Mackey
Zoom link: https://uwmadison.zoom.us/j/97034682490?pwd=Y082L0RSMUtQMnhPVGtyTmVGQ0thZz09
Title: Advances in Distribution Compression
Abstract: This talk will introduce three new tools for summarizing a probability distribution more effectively than independent sampling or standard Markov chain Monte Carlo thinning:
1. Given an initial n point summary (for example, from independent sampling or a Markov chain), kernel thinning finds a subset of only square-root n points with comparable worst-case integration error across a reproducing kernel Hilbert space.
2. If the initial summary suffers from biases due to off-target sampling, tempering, or burn-in, Stein thinning simultaneously compresses the summary and improves the accuracy by correcting for these biases.
3. Finally, Compress++ converts any unbiased quadratic-time thinning algorithm into a near-linear-time algorithm with comparable error.
These tools are especially well-suited for tasks that incur substantial downstream computation costs per summary point like organ and tissue modeling in which each simulation consumes 1000s of CPU hours.