Result Table Schema
Column | Data Type | Description |
---|---|---|
fold_num | INTEGER | Row 1: best_score (lowest test error) |
Row 2: average_score | ||
Row 3 and greater:: Fold number in [1, k]. | ||
validation_metric | DOUBLE PRECISION | Error on validation set of fold, where metric is mse, accuracy. or auroc, depending on Metric option. |
training_metric | DOUBLE PRECISION | [Appears only with EvaluateTraining ('true').] Error on training set of fold, where metric is mse, accuracy. or auroc, depending on Metric option. |
With Metric ('auroc'), both validation_metric and training_metric have the value NaN. When a row contains Nan, the function ignores it when calculating average_score.
OutputTable Schema
The OutputTable has the same schema as the GLML1L2 Output table. This table is the model table with the lowest test error among the k models..