Gavin Weiguang Ding

None

Research Team Lead

MASc Engineering Science, Simon Fraser University

I am a Machine Learning Researcher working at Borealis AI at Toronto, ON, Canada. I am interested in a wide range of topics of related to machine learning and artificial intelligence. Currently I spend the majority of my time researching adversarial robustness of deep learning models.

Before joining Borealis AI, I was an Artificial Intelligence Researcher at Kindred AI , where I worked on robotics and machine learning.

I was a research scientist (January 2014 to April 2015) working with Dr. Graham Taylor in School of Engineering, University of Guelph, Canada. There, I worked on deep learning and representation learning.

I got my Master of Applied Science degree from School of Engineering Science at Simon Fraser University (Burnaby, BC, Canada). I worked on medical image analysis with Dr. Mirza Faisal Beg. At SFU, I worked on 2 projects: 1) analysis of spatial-temporal zebrafish heart optical mapping data; 2) automated detection of retinal fluids from OCT images.

I got my Bachelor of Engineering degree from School of Advanced Engineering (academic talent program) at Beihang University (Beijing, China), where I studied Automation, specifically, flight control system.

Publications

June 3, 2020 On Minimax Optimality of GANs for Robust Mean Estimation
Twenty Third International Conference on Artificial Intelligence and Statistics (AISTATS), 2020
Authors: G. W. Ding , R. Huang, Y. Yu, K. Wu
April 30, 2020 Max-Margin Adversarial Training: Direct Input Space Margin Maximization through Adversarial Training
International Conference on Learning Representations (ICLR), 2020
Authors: G. W. Ding , Y. Sharma, K. Lui , R. Huang
Aug. 10, 2019 On the Effectiveness of Low Frequency Perturbations
International Joint Conference on Artificial Intelligence (IJCAI), 2019
Authors: Y. Sharma, G. W. Ding , M. Brubaker
May 6, 2019 On the Sensitivity of Adversarial Robustness to Input Data Distributions
International Conference on Learning Representations (ICLR), 2019
Authors: G. Weiguang Ding , K. Lui , T. Jin, L. Wang, R. Huang
Dec. 3, 2018 Dimensionality Reduction has Quantifiable Imperfections: Two Geometric Bounds
Neural Information Processing Systems (NeurIPS), 2018
Authors: K. Lui , G. W. Ding , R. Huang, R.J. McCann
April 30, 2018 Improving GAN Training via Binarized Representation Entropy (BRE) Regularization
International Conference on Learning Representations (ICLR), 2018
Authors: Y. Cao , G. W. Ding , K. Lui , R. Huang
None