Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2406.17433 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915602970443776 |
|---|---|
| author | Schrouff, Jessica Bellot, Alexis Rannen-Triki, Amal Malek, Alan Albuquerque, Isabela Gretton, Arthur D'Amour, Alexander Chiappa, Silvia |
| author_facet | Schrouff, Jessica Bellot, Alexis Rannen-Triki, Amal Malek, Alan Albuquerque, Isabela Gretton, Arthur D'Amour, Alexander Chiappa, Silvia |
| contents | Failures of fairness or robustness in machine learning predictive settings can be due to undesired dependencies between covariates, outcomes and auxiliary factors of variation. A common strategy to mitigate these failures is data balancing, which attempts to remove those undesired dependencies. In this work, we define conditions on the training distribution for data balancing to lead to fair or robust models. Our results display that, in many cases, the balanced distribution does not correspond to selectively removing the undesired dependencies in a causal graph of the task, leading to multiple failure modes and even interference with other mitigation techniques such as regularization. Overall, our results highlight the importance of taking the causal graph into account before performing data balancing. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_17433 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Mind the Graph When Balancing Data for Fairness or Robustness Schrouff, Jessica Bellot, Alexis Rannen-Triki, Amal Malek, Alan Albuquerque, Isabela Gretton, Arthur D'Amour, Alexander Chiappa, Silvia Machine Learning Failures of fairness or robustness in machine learning predictive settings can be due to undesired dependencies between covariates, outcomes and auxiliary factors of variation. A common strategy to mitigate these failures is data balancing, which attempts to remove those undesired dependencies. In this work, we define conditions on the training distribution for data balancing to lead to fair or robust models. Our results display that, in many cases, the balanced distribution does not correspond to selectively removing the undesired dependencies in a causal graph of the task, leading to multiple failure modes and even interference with other mitigation techniques such as regularization. Overall, our results highlight the importance of taking the causal graph into account before performing data balancing. |
| title | Mind the Graph When Balancing Data for Fairness or Robustness |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2406.17433 |