The performance of any binary classifier that provides a probabilistic output (rather than a binary value) can be visualized by splitting the target validation population into quantiles of a fixed count (usually 10, that is, deciles) and comparing the empirical incidence of the dependent variable to the probabilistic response outputted by the classifier.
The validationplot package makes this endeavour straightforward,
especially in conjunction with the tundra
package.
The package was originally authored by physicist Ryland Ely.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.