DM_getBestFit

This function returns a table that contains the algorithm names of the best fitting forecast for the given time series. This decision is based on the empirically determined standard error interval for the forecasts. If the best fitting algorithm cannot be calculated, NULL is returned instead of the corresponding sub-model.

Syntax

Function syntax


>>-DM_getBestFit--(--regModel--)-------------------------------><

Parameters

regModel: A value of type DM_RegressionModel

Return Value

The return value is of type TABLE

Table 1. Result Values for DM_getBestFit
Column	Data type	Purpose
submodel	VARCHAR(128)	name of the sub-model that is used to calculate the prediction. Each PMML document can contain multiple time series models for different columns of a given table.
bestFit	VARCHAR(30)	name of the algorithm with the best fitting forecast. Valid values are 'ARIMA', 'ExponentialSmoothing', or 'SeasonalTrendDecomposition'. If the best fitting algorithm cannot be calculated, NULL is returned instead of the corresponding sub-model.