ML
NeurIPS

Optimal Transport Model Distributional Robustness

October 4, 2023

Distributional robustness is a promising framework for training deep learning models that are less vulnerable to adversarial examples and data distribution shifts. Previous works have mainly focused on exploiting distributional robustness in the data space. In this work, we explore an optimal transport-based distributional robustness framework in model spaces. Specifically, we examine a model distribution within a Wasserstein ball centered on a given model distribution that maximizes the loss. We have developed theories that enable us to learn the optimal robust center model distribution. Interestingly, our developed theories allow us to flexibly incorporate the concept of sharpness awareness into training, whether it’s a single model, ensemble models, or Bayesian Neural Networks, by considering specific forms of the center model distribution. These forms include a Dirac delta distribution over a single model, a uniform distribution over several models, and a general Bayesian Neural Network. Furthermore, we demonstrate that Sharpness-Aware Minimization (SAM) is a specific case of our framework when using a Dirac delta distribution over a single model, while our framework can be seen as a probabilistic extension of SAM. To validate the effectiveness of our framework in the aforementioned settings, we conducted extensive experiments, and the results reveal remarkable improvements compared to the baselines.

Overall

2 minutes

Van-Anh Nguyen, Trung Le, Anh Tuan Bui, Thanh-Toan Do, Dinh Phung

NeurIPS 2023

Share Article

Related publications

ML
ICML Top Tier
May 16, 2024

Vy Vo, He Zhao, Trung Le, Edwin V. Bonilla, Dinh Phung

ML
ICML Top Tier
May 16, 2024

Vy Vo, Trung Le, Tung-Long Vuong, He Zhao, Edwin V. Bonilla, Dinh Phung

ML
ICML Top Tier
May 14, 2024

Ngoc Bui, Hieu Trung Nguyen, Viet Anh Nguyen, Rex Ying

GenAI
ML
ICML Top Tier
May 14, 2024

Bao Nguyen, Binh Nguyen, Hieu Nguyen, Viet Anh Nguyen