Listar MAT_Articles por fuente "BMC Bioinformatics (2017) 18:230"
Mostrando ítems 1-1 de 1
-
Intervention in prediction measure: a new approach to assessing variable importance for random forests
Springer Verlag (2017-05)Background Random forests are a popular method in many fields since they can be successfully applied to complex data, with a small sample size, complex interactions and correlations, mixed type predictors, etc. Furthermore, ...