English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Machine learning (ML) approaches to data analysis are now widely adopted inmany fields including epidemiology and medicine. To apply these approaches,confounds must first be removed as is commonly done by featurewise removal oftheir variance by linear regression before applying ML. Here, we show thiscommon approach to confound removal biases ML models, leading to misleadingresults. Specifically, this common deconfounding approach can leak informationsuch that what are null or moderate effects become amplified to near-perfectprediction when nonlinear ML approaches are subsequently applied. We identifyand evaluate possible mechanisms for such confound-leakage and providepractical guidance to mitigate its negative impact. We demonstrate thereal-world importance of confound-leakage by analyzing a clinical dataset whereaccuracy is overestimated for predicting attention deficit hyperactivitydisorder (ADHD) with depression as a confound. Our results have wide-reachingimplications for implementation and deployment of ML workflows and beg cautionagainst na\"ive use of standard confound removal approaches.

Confound-leakage: Confound Removal in Machine Learning Leads to Leakage