Kernel Thinning and Stein Thinning

Kernel Thinning and Stein Thinning

Explore kernel thinning and Stein thinning in this hour-long seminar presented by Lester Mackey from Microsoft and Stanford University. Delve into topics such as computational cardiology, inferential modeling, and distribution compression strategies. Learn about measuring approximation error, square root kernels, and the KernelSelfbalancing Hilbert Walk. Examine related work, practical applications, and results, including discussions on off-target sampling and kernel Stein discrepancy. Gain insights into Stein thinning terms and participate in a Q&A session to deepen your understanding of these advanced machine learning concepts.