This model was evaluated using {{ $ctrl.modelData.splitDesc.params.nFolds || 'K' }}-fold cross-test.

“Optimal” cut was found by optimizing for .

Be aware that some of these metrics (like accuracy) can be misleading since they assume that the target variable is balanced across your sample.

Partitioned model metrics

Keep in mind that metrics for the overall partitioned model are aggregates.
They have been computed from optimal-threshold results when available.
For others, like AUC, a weighted average has been used instead.

Averaging one-vs-all metrics across classes

One-vs-all metrics are computed for each class, the overall metric can be computed using the following methods.

  • Unweighted class average: all classes have equal weight. Better suited for classes of equal importance.
  • Weighted class average: classes are weighted by the number of rows for each class, or the sum of their sample weights if sample weights are specified.