Inflated prediction accuracy of neuropsychiatric biomarkers caused by data leakage in feature selection

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 machine learn...

Descripción completa

Detalles Bibliográficos
Autores principales: Shim, Miseon, Lee, Seung-Hwan, Hwang, Han-Jeong
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Nature Publishing Group UK 2021
Materias:
Article
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8042090/
https://www.ncbi.nlm.nih.gov/pubmed/33846489
http://dx.doi.org/10.1038/s41598-021-87157-3

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8042090/
https://www.ncbi.nlm.nih.gov/pubmed/33846489
http://dx.doi.org/10.1038/s41598-021-87157-3

Ejemplares similares