Re: [theory students] CS Theory Seminar this Wed, Nov 5: Liren Shan "Computing Small Volume Confidence Set for Arbitrary Distributions"

[Sandeep Silwal <silwal@xxxxxxxxxxx>]

Happening now! See you in 3610 or join on zoom here: https://uwmadison.zoom.us/j/8090370983

On Nov 5, 2025, at 2:02âPM, Sandeep Silwal <silwal@xxxxxxxxxxx> wrote:

Happening soon! 

On Nov 5, 2025, at 9:34âAM, Sandeep Silwal <sbsilwal@xxxxxxxx> wrote:

Seminar today! See you there:

Details: Wed Nov 5 at 2:30-3:30pm, Morgridge Hall  3610 

Title: 

Computing Small Volume Confidence Set for Arbitrary Distributions 

Abstract: We study the problem of learning a high-density region of an arbitrary distribution. Given a target coverage level, and sample access to an arbitrary distribution D, we want to output a confidence set S such that S achieves the desired coverage and the volume of S is as small as possible. We show that this problem is statistically intractable in the most general setting. Then, we restrict our attention to competing with sets from a set family C with bounded VC dimension and provide approximation algorithms to compute a small volume set with desired coverage. Finally, we demonstrate the application of our results to conformal prediction.

Sandeep

On Nov 4, 2025, at 9:46âPM, Sandeep Silwal <silwal@xxxxxxxxxxx> wrote:

Hi everyone!

We have the seminar tomorrow, see you there! There are still some meeting spots with Liren if you want to chat with him. Please sigh up here: https://docs.google.com/spreadsheets/d/1yyB6znjmxrTtMgpTBhOhX-HDdsrBxaDNh8U3gBviFio/edit?gid=0#gid=0

On Nov 3, 2025, at 9:19âAM, Sandeep Silwal <silwal@xxxxxxxxxxx> wrote:

This week (Wed Nov 5) we have a visitor Liren Shan (https://lirenshan.github.io/)  from TTIC! If you would like to meet with Liren, please sign up here: https://docs.google.com/spreadsheets/d/1yyB6znjmxrTtMgpTBhOhX-HDdsrBxaDNh8U3gBviFio/edit?usp=sharing

Details: Wed Nov 5 at 2:30-3:30pm, Morgridge Hall  3610 

Title: 

Computing Small Volume Confidence Set for Arbitrary Distributions 

Abstract: We study the problem of learning a high-density region of an arbitrary distribution. Given a target coverage level, and sample access to an arbitrary distribution D, we want to output a confidence set S such that S achieves the desired coverage and the volume of S is as small as possible. We show that this problem is statistically intractable in the most general setting. Then, we restrict our attention to competing with sets from a set family C with bounded VC dimension and provide approximation algorithms to compute a small volume set with desired coverage. Finally, we demonstrate the application of our results to conformal prediction.

See you there!

Sandeep