Saved in:
Bibliographic Details
Main Authors: Wegel, Tobias, Kovačević, Filip, Ţifrea, Alexandru, Yang, Fanny
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.08849
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Simultaneously addressing multiple objectives is becoming increasingly important in modern machine learning. At the same time, data is often high-dimensional and costly to label. For a single objective such as prediction risk, conventional regularization techniques are known to improve generalization when the data exhibits low-dimensional structure like sparsity. However, it is largely unexplored how to leverage this structure in the context of multi-objective learning (MOL) with multiple competing objectives. In this work, we discuss how the application of vanilla regularization approaches can fail, and propose a two-stage MOL framework that can successfully leverage low-dimensional structure. We demonstrate its effectiveness experimentally for multi-distribution learning and fairness-risk trade-offs.