comp_class_stats: Accuracy, assignment probability, Brier scores and ROC for...

In PiotrTymoszuk/caretExtra: Extra functionality for predictions and quality control of caret models

Accuracy, assignment probability, Brier scores and ROC for outcome classes.

Description

Those internally utilized functions compute:

• comp_class_stats(): average class assignment probability and Brier scores for the outcome classes.

• comp_class_roc(): accuracy, sensitivity, specificity, recall etc. for each of the class compared with the rest.

Usage

comp_class_stats(predx_object)

comp_class_roc(predx_object)

Arguments

predx_object

a predx class object.

Details

For regression, NULL is returned with a warning. comp_class_roc() uses multiClassSummary.

Value

a data frame with columns representing the requested statistics for particular classes.

PiotrTymoszuk/caretExtra documentation built on Oct. 15, 2023, 10:03 p.m.