Kaiwen Wu

MMath student

David R. Cheriton School of Computer Science

University of Waterloo

Email: kaiwen.wu@uwaterloo.ca

GitHub   |   Twitter   |   Scholar

About Me

I recently completed my master degree in computer science at the University of Waterloo under the supervision of Prof. Yaoliang Yu. I wrote my thesis on optimization algorithms for Wasserstein adversarial robustness. Previously, I did my undergraduate in computer science at Nanjing University.

My research is in machine learning and optimization. I am interested in developing efficient and robust learning algorithms. In particular, I am interested in large-scale optimization, adversarial robustness, generative models.

Newton-type Methods for Minimax Optimization
Guojun Zhang, Kaiwen Wu, Pascal Poupart, Yaoliang Yu
Preprint [arXiv]

Stronger and Faster Wasserstein Adversarial Attacks
Kaiwen Wu, Allen Houze Wang and Yaoliang Yu
The 37th International Conference on Machine Learning (ICML 2020)
paper] [arXiv] [code]

On Minimax Optimality of GANs for Robust Mean Estimation
Kaiwen Wu, Gavin Weiguang Ding, Ruitong Huang and Yaoliang Yu
The 23rd International Conference on Artificial Intelligence and Statistics (AISTATS 2020)
paper] [code]

Understanding Adversarial Robustness: The Trade-off between Minimum and Average Margin
Kaiwen Wu and Yaoliang Yu
NeurIPS 19 Workshop on Maching Learning with Guarantees
arXiv][workshop version]

Kaiwen Wu, Wasserstein Adversarial Robustness. MMath Thesis 2020. [pdf]

On Minimax Optimality of GANs for Robust Mean Estimation [slides]
• AISTATS, Aug 2020

Stronger and Faster Wasserstein Adversarial Attacks [slides]
• Vector Institute, Sep 2020
• ICML, Jul 2020

Wasserstein Adversarial Examples [slides]
• Group Seminar, Summer 2019

L1 Norm Projection