# Distributionally Robust Group Backwards Compatibility

Source: [Distributionally Robust Group Backwards Compatibility](<https://www.sota2.com/research/paper/68fa6b7f1e209131cea2e9f3>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:56:51.287Z

Authors: Martin Bertran, Natalia Martinez, Alex Oesterling, Guillermo Sapiro

Published: 2021-12-20T00:30:26.000Z

## Abstract

Machine learning models are updated as new data is acquired or new architectures are developed. These updates usually increase model performance, but may introduce backward compatibility errors, where individual users or groups of users see their performance on the updated model adversely affected. This problem can also be present when training datasets do not accurately reflect overall population demographics, with some groups having overall lower participation in the data collection process, posing a significant fairness concern. We analyze how ideas from distributional robustness and minimax fairness can aid backward compatibility in this scenario, and propose two methods to directly address this issue. Our theoretical analysis is backed by experimental results on CIFAR-10, CelebA, and Waterbirds, three standard image classification datasets. Code available at github.com/natalialmg/GroupBC

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2112.10290>)

[PDF](<https://arxiv.org/pdf/2112.10290.pdf>)