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...

Description complète

Enregistré dans:
Détails bibliographiques
Auteurs principaux: Miseon Shim, Seung-Hwan Lee, Han-Jeong Hwang
Format: article
Langue:EN
Publié: Nature Portfolio 2021
Sujets:
R
Q
Accès en ligne:https://doaj.org/article/ad61d8cc3ca4460d809b2864f5cb0832
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!