Diagnostic measures on models

Diagnostic measures help you reliably assess the quality of predictive models.

Usage of diagnostic measures
By using the provided stored procedures, you can, for example, build a confusion matrix for classification predictions to calculate different rates, errors, or other factors for the confusion matrix or for the prediction models.
Functions for diagnostic measures
The functions for diagnostic measures are implemented in several stored procedures for different calculations. For example, you can use the ACC stored procedure to calculate the classification accuracy, or the RAE stored procedure to calculate the relative absolute error of regression predictions.

Parent topic: Analytic stored procedures