Learning Deep Representations with Optimal Transport

The next statistics seminar is presented by Dr. He Zhao from Data61 at CSIRO. The details of the talk are given below.

Title: Learning Deep Representations with Optimal Transport
Speaker: He Zhao, Data61, CSIRO
Time and location: 2pm-3pm 8 March 2024 at Chemistry Lecture Theatre 4, Room 236
Abstract: Originated from the works of mathematicians, statisticians, and economists, Optimal Transport (OT) is a powerful tool for resource allocation. Recently, OT has gained significant attention and utility in machine learning and deep learning, particularly in areas where the comparison of probability measures is essential. In this talk, I will introduce two recent works of mine on applying OT for deep representation learning that captures essential structural information in the data, leading to improved generalisation and robustness. One is on the task of image data augmentation for imbalanced problems and the other is on missing value imputation.
Bio: He Zhao is a Senior Research Scientist in machine learning at CSIRO’s Data61. Before joining Data61, he received his PhD from Monash University, working on Bayesian modelling and inference and then worked as a research follow, working on optimal transport applications. His current research interest is developing probabilistic approaches that improve robustness, generalisation, uncertainty estimation, and interpretation of machine/deep learning. He has published 35+ research papers in machine learning and related venues ranked as CORE A* or A, among which, 20 papers were in NeurIPS, ICML, ICLR, and AISTATS, the leading machine learning conferences. His paper in KDD 2023 (the best venue of data mining and machine learning applications) won the best student paper award (1 out of 1416). He is on the Editorial board of Springer Machine Learning Journal. His research website is at https://hezgit.github.io.