SDS Statistics Seminar Series

Topic

Randomized Tensor-network Algorithms for Random Data in High-dimensions

Speaker

Yuehaw KHOO, Assistant Professor, University of Chicago

Host

Fangda SONG, Assistant Professor, School of Data Science, CUHK-Shenzhen

Date

13 March (Thursday), 2025

Time

4:00 PM- 5:00 PM, Beijing Time

Format

Hybrid

Venue

Room 401, Dao Yuan Building

Zoom Link

https://cuhk-edu-cn.zoom.us/j/94583716283?pwd=IyExqTaai0ctadaBTPzfu8P9nbvYxJ.1

Meeting ID: 945 8371 6283, Password: 910154

Language

English

Abstract

Tensor-networks have long been employed to solve the high-dimensional Schrödinger equation deterministically, demonstrating linear complexity scaling with respect to dimensionality. Recently, this ansatz has found applications in various machine learning scenarios, including supervised learning and generative modeling, where the data originates from a random process. In this talk, we present a new perspective on randomized linear algebra, showcasing its usage in estimating a density as a tensor-network from i.i.d. samples of a distribution, without the curse of dimensionality, and without the use of optimization techniques. Moreover, we illustrate how this concept can combine the strengths of Monte-Carlo and tensor-network methods for solving high-dimensional PDEs, resulting in enhanced flexibility for both approaches.

Biography

Yuehaw Khoo is an Assistant Professor at the University of Chicago, in the Committee on Computational and Applied Mathematics under the Department of Statistics. He works in computational mathematics, focusing on solving high-dimensional problems in physics and biology using optimization, machine learning and tensor methods. He completed his postdoctoral studies at Stanford University and his graduate studies at Princeton University. He has received NSF Career Award and Sloan Research Fellowship.