Robust and Sparse Interpretation of Deep Neural Networks

Speaker: Dr. Farzan Farnia
         Department of Computer Science and Engineering
         The Chinese University of Hong Kong

Title:  "Robust and Sparse Interpretation of Deep Neural Networks"

Date:   Monday, 13 March 2023

Time:   4:00pm - 5:00pm

Venue:  Lecture Theater F
        (Leung Yat Sing Lecture Theater), near lift 25/26
        HKUST

Abstract:

Explaining the predictions of deep neural networks has been a topic of
great interest in the machine learning community. In this talk, we will
focus on gradient-based saliency maps for interpreting neural network
models. We will discuss the vulnerability of standard gradient-based
interpretation schemes to input perturbations, and introduce MoreauGrad as
an interpretation scheme based on a neural network's Moreau envelope. We
prove the certifiable robustness of MoreauGrad to norm-bounded input
perturbations, and subsequently propose a sparse version of MoreauGrad by
applying L1-norm regularization to its formulation. We discuss the
robustness properties of Sparse MoreauGrad and display the visual
performance of our proposed interpretation scheme in application to
standard image datasets.


******************
Biography:

Farzan Farnia is an Assistant Professor of Computer Science and
Engineering at The Chinese University of Hong Kong. Prior to joining CUHK,
he was a postdoctoral research associate at the Laboratory for Information
and Decision Systems, Massachusetts Institute of Technology, from
2019-2021. He received his master's and PhD degrees in electrical
engineering from Stanford University and his bachelor's degrees in
electrical engineering and mathematics from Sharif University of
Technology. At Stanford, he was a graduate research assistant at the
Information Systems Laboratory advised by David Tse. Farzan's research
interests span statistical learning theory, information theory, and convex
optimization.