Hamdan, Sami
[Author];
Love, Bradley C.
[Author];
Polier, Georg von
[Author];
Weis, Susanne
[Author];
Schwender, Holger
[Author];
Eickhoff, Simon
[Author];
Patil, Kaustubh
[Author]
Cofound-Leakage: Confound Removal In Machine Learning Leads To Leakage
You can manage bookmarks using lists, please log in to your user account for this.
Media type:
Electronic Conference Proceeding
Title:
Cofound-Leakage: Confound Removal In Machine Learning Leads To Leakage
Contributor:
Hamdan, Sami
[Author];
Love, Bradley C.
[Author];
Polier, Georg von
[Author];
Weis, Susanne
[Author];
Schwender, Holger
[Author];
Eickhoff, Simon
[Author];
Patil, Kaustubh
[Author]
Published in:doi:10.34734/FZJ-2023-03045 ; Organization for Human Brain Mapping (OHBM), Montreal, Canada, 2023-07-22 - 2023-07-26
Language:
English
DOI:
https://doi.org/10.34734/FZJ-2023-03045
Origination:
Footnote:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Description:
Modern Machine Learning (ML) approaches are now regularly employed forindividual-level prediction, e.g. personalized medicine.Particularly in such critical-decision making, it is of utmost importance to not onlyachieve high accuracy but also to trust that models rely on actual features-targetrelationships [1, 2]. To this end, it is crucial to consider confounding variables as theycan obstruct the features-target relationship. For instance, a researcher might wantto identify a biomarker showing high classification accuracy between controls andpatients. However, the model might have just learned simpler confounders like ageor sex as a good proxy of the disease [3]. To counteract such unwanted confoundingeffects, investigators often use linear models to remove confounding variables fromeach feature separately before employing ML. While this confound regression (CR)approach is popular [4], its pitfalls, especially when paired with non-linear MLmodels, are not well understood.