---
OA_place: repository
OA_type: green
_id: '20300'
abstract:
- lang: eng
  text: 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.
acknowledgement: "We thank Junhyung Park for valuable feedback on the manuscript.
  AT was supported by a PhD fellowship from the Swiss Data Science Center. TW was
  supported by the SNF Grant 204439. This work was done in part while TW and FY were
  visiting the Simons Institute for the Theory of\r\nComputing."
alternative_title:
- PMLR
article_processing_charge: No
arxiv: 1
author:
- first_name: Tobias
  full_name: Wegel, Tobias
  last_name: Wegel
- first_name: Filip
  full_name: Kovačević, Filip
  id: d0258e7b-50b8-11ef-ad56-8b9f537b6b1b
  last_name: Kovačević
- first_name: Alexandru
  full_name: Ţifrea, Alexandru
  last_name: Ţifrea
- first_name: Fanny
  full_name: Yang, Fanny
  last_name: Yang
citation:
  ama: 'Wegel T, Kovačević F, Ţifrea A, Yang F. Learning Pareto manifolds in high
    dimensions: How can regularization help? In: <i>The 28th International Conference
    on Artificial Intelligence and Statistics</i>. Vol 258. ML Research Press; 2025:4591-4599.'
  apa: 'Wegel, T., Kovačević, F., Ţifrea, A., &#38; Yang, F. (2025). Learning Pareto
    manifolds in high dimensions: How can regularization help? In <i>The 28th International
    Conference on Artificial Intelligence and Statistics</i> (Vol. 258, pp. 4591–4599).
    Mai Khao, Thailand: ML Research Press.'
  chicago: 'Wegel, Tobias, Filip Kovačević, Alexandru Ţifrea, and Fanny Yang. “Learning
    Pareto Manifolds in High Dimensions: How Can Regularization Help?” In <i>The 28th
    International Conference on Artificial Intelligence and Statistics</i>, 258:4591–99.
    ML Research Press, 2025.'
  ieee: 'T. Wegel, F. Kovačević, A. Ţifrea, and F. Yang, “Learning Pareto manifolds
    in high dimensions: How can regularization help?,” in <i>The 28th International
    Conference on Artificial Intelligence and Statistics</i>, Mai Khao, Thailand,
    2025, vol. 258, pp. 4591–4599.'
  ista: 'Wegel T, Kovačević F, Ţifrea A, Yang F. 2025. Learning Pareto manifolds in
    high dimensions: How can regularization help? The 28th International Conference
    on Artificial Intelligence and Statistics. AISTATS: Conference on Artificial Intelligence
    and Statistics, PMLR, vol. 258, 4591–4599.'
  mla: 'Wegel, Tobias, et al. “Learning Pareto Manifolds in High Dimensions: How Can
    Regularization Help?” <i>The 28th International Conference on Artificial Intelligence
    and Statistics</i>, vol. 258, ML Research Press, 2025, pp. 4591–99.'
  short: T. Wegel, F. Kovačević, A. Ţifrea, F. Yang, in:, The 28th International Conference
    on Artificial Intelligence and Statistics, ML Research Press, 2025, pp. 4591–4599.
conference:
  end_date: 2025-05-05
  location: Mai Khao, Thailand
  name: 'AISTATS: Conference on Artificial Intelligence and Statistics'
  start_date: 2025-05-03
date_created: 2025-09-07T22:01:35Z
date_published: 2025-05-01T00:00:00Z
date_updated: 2025-09-09T07:00:34Z
day: '01'
department:
- _id: MaMo
external_id:
  arxiv:
  - '2503.08849'
intvolume: '       258'
language:
- iso: eng
main_file_link:
- open_access: '1'
  url: https://doi.org/10.48550/arXiv.2503.08849
month: '05'
oa: 1
oa_version: Preprint
page: 4591-4599
publication: The 28th International Conference on Artificial Intelligence and Statistics
publication_identifier:
  eissn:
  - 2640-3498
publication_status: published
publisher: ML Research Press
quality_controlled: '1'
scopus_import: '1'
status: public
title: 'Learning Pareto manifolds in high dimensions: How can regularization help?'
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 258
year: '2025'
...