Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 218 - Leveraging and Advancing Deep Learning Techniques in Biomedical Related Fields
Type: Invited
Date/Time: Wednesday, August 11, 2021 : 10:00 AM to 11:50 AM
Sponsor: International Chinese Statistical Association
Abstract #316665
Title: Efficient Multi-Modal Sampling via Tempered Distribution Flow
Author(s): Xiao Wang* and Yixuan Qiu
Companies: Purdue University and Shanghai University of Finance and Economics
Keywords: deep neural network; generative models; MCMC; normalization flow; sampling
Abstract:

Sampling from high-dimensional distributions is a fundamental problem in statistical research and practice, and has become a central task in many machine learning models such as energy-based models and deep generative models. However, great challenges emerge when the density function contains multiple modes that are isolated with each other. We tackle this difficulty by fitting an invertible transformation function based on normalizing flow techniques, such that the original distribution is warped into a new one that is much easier to sample from. To address the multi-modality issue, our method adaptively learns a sequence of tempered distributions, which we term as a tempered distribution flow, to progressively approach the desired distribution. Numerical experiments demonstrate the superior performance of this novel sampler compared to traditional methods.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program