Exploring the variable importance in random forests under correlations: a general concept applied to donor organ quality in post-transplant survival

Random Forests are a powerful and frequently applied Machine Learning tool. The permutation variable importance (VIMP) has been proposed to improve the explainability of such a pure prediction model. It describes the expected increase in prediction error after randomly permuting a variable and distu...

Descripción completa

Detalles Bibliográficos
Autores principales: Wies, Christoph, Miltenberger, Robert, Grieser, Gunter, Jahn-Eimermacher, Antje
Formato: Online Artículo Texto
Lenguaje:English
Publicado: BioMed Central 2023
Materias:
Research
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10507897/
https://www.ncbi.nlm.nih.gov/pubmed/37726680
http://dx.doi.org/10.1186/s12874-023-02023-2

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10507897/
https://www.ncbi.nlm.nih.gov/pubmed/37726680
http://dx.doi.org/10.1186/s12874-023-02023-2

Ejemplares similares