7.00.02 - Output - Aster Analytics

Teradata Aster® Analytics Foundation User GuideUpdate 2

Product
Aster Analytics
Release Number
7.00.02
Published
September 2017
Content Type
Programming Reference
User Guide
Publication ID
B700-1022-700K
Language
English (United States)
Last Update
2018-04-17

The ConfusionMatrix function returns a success message and creates 3 output tables:

  • output_table_1, a confusion matrix (also called a contingency table)
  • output_table_2, which contains overall statistics
  • output_table_3, which contains statistics for each class
    ConfusionMatrix Output Table 1 (Confusion Matrix) Schema
    Column Name Data Type Description
    observation/predict VARCHAR One row for each unique value of observed_column in the input table.
    predicted_class INTEGER The number of times that items in the observed_column were classified as predicted_class.
    ConfusionMatrix Output Table 2 (Overall Statistics) Schema
    Column Name Data Type Description
    key VARCHAR Each row contains one of the following statistic names:
    • Accuracy
    • 95% CI
    • Null Error Rate
    • P-Value [Acc > NIR]
    • Kappa
    • McNemar Test P-Value
    value DOUBLE PRECISION Values of the statistics.

The schema of output_table_3 depends on the number of classes.

ConfusionMatrix Output Table 3 (Class Statistics) Schema for Two Classes
Column Name Data Type Description
key VARCHAR Each row contains one of the following statistic names:
  • Sensitivity
  • Specificity
  • Pos Pred Value
  • Neg Pred Value
  • Prevalence
  • Detection Rate
  • Detection Prevalence
  • Balanced Accuracy
value DOUBLE PRECISION Values of the statistics.
ConfusionMatrix Output Table 3 (Class Statistics) Schema for More than Two Classes
Column Name Data Type Description
key VARCHAR Each row contains one of the following statistic names:
  • Sensitivity
  • Specificity
  • Pos Pred Value
  • Neg Pred Value
  • Prevalence
  • Detection Rate
  • Detection Prevalence
  • Balanced Accuracy
class:expect_class DOUBLE PRECISION Values of the statistics for the class expect_class.

If you specify the Classes argument, there is one column for each specified value. Otherwise, there is one column for each unique value in the observed column of the input table.