Learning Under Adversarial and Interventional Shifts

03/29/2021
by   Harvineet Singh, et al.
0

Machine learning models are often trained on data from one distribution and deployed on others. So it becomes important to design models that are robust to distribution shifts. Most of the existing work focuses on optimizing for either adversarial shifts or interventional shifts. Adversarial methods lack expressivity in representing plausible shifts as they consider shifts to joint distributions in the data. Interventional methods allow more expressivity but provide robustness to unbounded shifts, resulting in overly conservative models. In this work, we combine the complementary strengths of the two approaches and propose a new formulation, RISe, for designing robust models against a set of distribution shifts that are at the intersection of adversarial and interventional shifts. We employ the distributionally robust optimization framework to optimize the resulting objective in both supervised and reinforcement learning settings. Extensive experimentation with synthetic and real world datasets from healthcare demonstrate the efficacy of the proposed approach.

READ FULL TEXT
research
10/21/2021

A Fine-Grained Analysis on Distribution Shift

Robustness to distribution shifts is critical for deploying machine lear...
research
10/12/2021

Tracking the risk of a deployed model and detecting harmful distribution shifts

When deployed in the real world, machine learning models inevitably enco...
research
01/28/2022

Certifying Model Accuracy under Distribution Shifts

Certified robustness in machine learning has primarily focused on advers...
research
05/24/2023

Non-adversarial Robustness of Deep Learning Methods for Computer Vision

Non-adversarial robustness, also known as natural robustness, is a prope...
research
09/18/2022

Towards Robust Off-Policy Evaluation via Human Inputs

Off-policy Evaluation (OPE) methods are crucial tools for evaluating pol...
research
02/06/2023

Bitrate-Constrained DRO: Beyond Worst Case Robustness To Unknown Group Shifts

Training machine learning models robust to distribution shifts is critic...
research
04/24/2023

Addressing distributional shifts in operations management: The case of order fulfillment in customized production

To meet order fulfillment targets, manufacturers seek to optimize produc...

Please sign up or login with your details

Forgot password? Click here to reset