Skip to content
#

feature-selection

Here are 332 public repositories matching this topic...

MLopez-Ibanez
MLopez-Ibanez commented Sep 16, 2018

In the table given in https://mlr-org.github.io/mlr/articles/tutorial/cost_sensitive_classif.html#binary-classification-problems
true are rows and pred are columns, however, the cost matrix at the top is defined as "The costs c(k,l) for predicting class k if the true label is l", so pred are rows and true are columns.

This causes some confusion in the example for binary classification.

In

You can’t perform that action at this time.