Inflated prediction accuracy of neuropsychiatric biomarkers caused by data leakage in feature selection
Abstract In recent years, machine learning techniques have been frequently applied to uncovering neuropsychiatric biomarkers with the aim of accurately diagnosing neuropsychiatric diseases and predicting treatment prognosis. However, many studies did not perform cross validation (CV) when using mach...
Enregistré dans:
Auteurs principaux: | , , |
---|---|
Format: | article |
Langue: | EN |
Publié: |
Nature Portfolio
2021
|
Sujets: | |
Accès en ligne: | https://doaj.org/article/ad61d8cc3ca4460d809b2864f5cb0832 |
Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|