29 citations · 158 across the 26 of their papers we have counts for
7 papers · 1 filter
Towards Understanding the Dynamics of the First-Order Adversaries
Zhun Deng, Hangfeng He, Jiaoyang Huang +1
An acknowledged weakness of neural networks is their vulnerability to adversarial perturbations to the inputs. To improve the robustness of these models, one of the most popular de…
Interpreting Robust Optimization via Adversarial Influence Functions
Zhun Deng, Cynthia Dwork, Jialiang Wang +1
Robust optimization has been widely used in nowadays data science, especially in adversarial training. However, little research has been done to quantify how robust optimization ch…
Toward Better Generalization Bounds with Locally Elastic Stability
Zhun Deng, Hangfeng He, Weijie J. Su
Algorithmic stability is a key characteristic to ensure the generalization ability of a learning algorithm. Among different notions of stability, \emph{uniform stability} is arguab…
How Does Mixup Help With Robustness and Generalization?
Linjun Zhang, Zhun Deng, Kenji Kawaguchi +2
Mixup is a popular data augmentation technique based on taking convex combinations of pairs of examples and their labels. This simple technique has been shown to substantially impr…
Decision-Aware Conditional GANs for Time Series Data
He Sun, Zhun Deng, Hui Chen +1
We introduce the decision-aware time-series conditional generative adversarial network (DAT-CGAN) as a method for time-series generation. The framework adopts a multi-Wasserstein l…
Representation via Representations: Domain Generalization via Adversarially Learned Invariant Representations
Zhun Deng, Frances Ding, Cynthia Dwork +4
We investigate the power of censoring techniques, first developed for learning {\em fair representations}, to address domain generalization. We examine {\em adversarial} censoring…
Improving Adversarial Robustness via Unlabeled Out-of-Domain Data
Zhun Deng, Linjun Zhang, Amirata Ghorbani +1
Data augmentation by incorporating cheap unlabeled data from multiple domains is a powerful way to improve prediction especially when there is limited labeled data. In this work, w…