ROC Output - Teradata Vantage

Machine Learning Engine Analytic Function Reference

Product
Teradata Vantage
Release Number
8.00
1.0
Published
May 2019
Language
English (United States)
Last Update
2019-11-22
dita:mapPath
blj1506016597986.ditamap
dita:ditavalPath
blj1506016597986.ditaval
dita:id
B700-4003
lifecycle
previous
Product Category
Teradata Vantageā„¢

The function outputs a message and a table. The output table schema depends on the ROCValues argument.

Output Message Schema

Column Data Type Description
info VARCHAR Reports whether function completed.

OutputTable Schema, ROCValues ('true') (Default)

The table has one row for each threshold for each model, and contains only ROC values.

Column Data Type Description
model VARCHAR [Appears only with ModelIDColumn argument.] Model identifier or partition for ROC curve associated with observation, taken from model_id_column.
threshold DOUBLE PRECISION Threshold at which function classifies an observation as positive.
tpr DOUBLE PRECISION True positive rate for threshold (number of observations correctly predicted as positive based on threshold, divided by number of observations known to be positive).
fpr DOUBLE PRECISION False positive rate for threshold (number of observations incorrectly predicted as positive based on threshold, divided by number of observations known to be negative).

OutputTable Schema, ROCValues ('false')

The table has the following:
  • One row for each model
  • No ROC values
  • AUC values, Gini values, or both (depending on AUC and Gini arguments)
Column Data Type Description
model VARCHAR [Appears only with ModelIDColumn argument.] Model identifier or partition for ROC curve associated with observation, taken from model_id_column.
AUC DOUBLE PRECISION Area under ROC curve for data in partition. With AUC ( 'false'), this value is NULL.
Gini DOUBLE PRECISION Gini coefficient for ROC curve for data in partition. With Gini ('false'), this value is NULL.