Workshops

Fall 2021

Sampling Algorithms and Geometries on Probability Distributions

Sep 27, 2021 to Oct 1, 2021

Return to event »

All talks are listed in Pacific Time. Please note some speakers may present remotely. Schedule subject to change.

Monday, September 27th, 2021
9:00 am – 9:20 am		Coffee & Check In
9:20 am – 9:30 am		Opening Remarks
9:30 am – 10:10 am		How Many Clusters - An Algorithmic Answer Ravi Kannan (Microsoft Research India)
10:10 am – 10:20 am		Break
10:20 am – 11:00 am		John's Walk and Coordinate Hit-and-Run Hariharan Narayanan (Tata Institute, Mumbai)
11:00 am – 11:30 am		Break
11:30 am – 12:10 pm		Faster Polytope Rounding, Sampling, and Volume Computation via a Sub-Linear Ball Walk Oren Mangoubi (WPI)
12:10 pm – 2:00 pm		Lunch
2:00 pm – 2:40 pm		Improved Dimension Dependence for the Metropolis-Adjusted Langevin Algorithm and Lower Bounds for Sampling Sinho Chewi (MIT)
2:40 pm – 2:50 pm		Break
2:50 pm – 3:30 pm		Lower Bounds on Metropolized Sampling Methods for Well-Conditioned Distributions Kevin Tian (Stanford)
3:30 pm – 4:00 pm		Break
4:00 pm – 4:40 pm		Mean-Square Analysis of the Mirror Langevin Algorithm Andre Wibisono (Yale)
4:40 pm – 5:30 pm		Reception

Tuesday, September 28th, 2021
9:00 am – 9:30 am		Coffee & Check In
9:30 am – 10:10 am		On Langevin Monte Carlo for Log-Concave Densities Arnak Dalalyan (ENSAE Paris)
10:10 am – 10:20 am		Break
10:20 am – 11:00 am		Quantitative Convergence of Unadjusted Langevin Monte Carlo and Application to Stochastic Approximation Alain Durmus (ENS Paris-Saclay)
11:00 am – 11:30 am		Break
11:30 am – 12:10 pm		Primal--Dual Optimization and Application to Sampling Adil Salim (KAUST)
12:10 pm – 2:00 pm		Lunch
2:00 pm – 2:40 pm		Structured Logconcave Sampling with a Restricted Gaussian Oracle Ruoqi Shen (University of Washington)
2:40 pm – 2:50 pm		Break
2:50 pm – 3:30 pm		Algorithms for Sampling Convex Bodies Aditi Laddha (Georgia Institute of Technology)
3:30 pm – 4:00 pm		Break
4:00 pm – 4:40 pm		Random Coordinate Descent and Langevin Monte Carlo Qin Li (UW Madison)

Wednesday, September 29th, 2021
9:00 am – 9:30 am		Coffee & Check In
9:30 am – 10:10 am		A Geometric and Analytical Perspective on Adversarial Learning Problems Nicolas Garcia Trillos (University of Wisconsin-Madison)
10:10 am – 10:20 am		Break
10:20 am – 11:00 am		Current and Future Applications of Sampling Algorithms in Modelling Biochemical Networks Ronan Fleming (Leiden University)
11:00 am – 11:30 am		Break
11:30 am – 12:10 pm		A Blob Method for Diffusion and Applications to Sampling and Two Layer Neural Networks Katy Craig (UC Santa Barbara)
12:10 pm – 2:00 pm		Lunch
2:00 pm – 2:40 pm		A Mean-Field Theory of Lazy Training in Two-Layer Neural Nets: Entropic Regularization and Controlled McKean-Vlasov Dynamics Belinda Tzen (UIUC)
2:40 pm – 2:50 pm		Break
2:50 pm – 3:30 pm		Kernel Stein Discrepancy Descent Anna Korba (ENSAE Paris)
3:30 pm – 4:00 pm		Break
4:00 pm – 4:50 pm		Open Problems Session

Thursday, September 30th, 2021
9:00 am – 9:30 am		Coffee & Check In
9:30 am – 10:10 am		Sampling Convex Bodies: A Status Report Santosh Vempala (Georgia Tech)
10:10 am – 10:20 am		Break
10:20 am – 11:00 am		Quantitative Convergence Analysis of Hypocoercive Sampling Dynamics Jianfeng Lu (Duke University)
11:00 am – 11:30 am		Break
11:30 am – 12:10 pm		Minimax Mixing Time of the Metropolis-Adjusted Langevin Algorithm for Log-Concave Sampling Yuansi Chen (Duke University)
12:10 pm – 2:00 pm		Lunch
2:00 pm – 2:40 pm		High-Order Langevin Diffusion Yields an Accelerated MCMC Algorithm Wenlong Mou (UC Berkeley)
2:40 pm – 2:50 pm		Break
2:50 pm – 3:30 pm		When Is the Convergence Time of Langevin Algorithms Dimension Independent: A Composite Optimization Viewpoint Yian Ma (UC San Diego)
3:30 pm – 4:00 pm		Break
4:00 pm – 4:40 pm		Generative Models for Sampling on Supports Inspired by Observed Data Susan Holmes (Stanford)

Friday, October 1st, 2021
9:00 am – 9:30 am		Coffee & Check In
9:30 am – 10:10 am		MCMC Sampling Around Redistricting and Gerrymandering Jonathan Mattingly (Duke University)
10:10 am – 10:20 am		Break
10:20 am – 11:00 am		Sliced Normalizing Flow Optimization and Monte Carlo Uros Seljak (UC Berkeley)
11:00 am – 11:30 am		Break
11:30 am – 12:10 pm		Adaptive Wavelet Distillation from Neural Networks through Interpretation Bin Yu (UC Berkeley)
12:10 pm – 2:00 pm		Lunch
2:00 pm – 2:40 pm		Quantitative Faber-Krahn Inequalities and the ACF Monotonicity Formula Robin Neumayer (Northwestern University)
2:40 pm – 2:50 pm		Break
2:50 pm – 3:30 pm		On the Convergence of Monte Carlo Methods with Stochastic Gradients Quanquan Gu (UCLA)
3:30 pm – 4:00 pm		Break
4:00 pm – 4:40 pm		Optimal Gradient-based Algorithms for Non-concave Bandit Optimization Qi Lei (Princeton)