Modifying Machine learning analyzer settings
When you want to change the decision threshold set for a Machine learning (ML) analyzer or disable the analyzer, you can modify the settings from the Analyzer page.
Before you begin
You must have completed the following tasks:
- Ensured that you are assigned a role as a Viewer, Tester, or Project Owner in the project.
- Ran the Performance Test or Schedule that matches the minimum requirements for ML analysis in Test Hub. See Configuring a run of a Compound Test that contains performance tests or Configuring a run of a Rate Schedule or VU Schedule.
- Viewed the insights into tests provided by ML analysis. See Viewing Machine learning insights.
- Viewed the reactions from members of your project to the ML insight. See Viewing reactions marked by members to Machine learning insights.
- Viewed the history of member reactions for a specific parameter analyzer from the Analyzers page. See Viewing Machine learning analyzers.
Procedure
-
Log in to Test Hub.
The team space that contains your project is displayed.
-
Click in the navigation pane.
The Analyzers page is displayed. You can view the parameters that are supported for ML analysis listed in a tabular format.
-
Identify the parameter whose analyzer you want to modify, and then click the
Edit Analyzer icon
in the Actions column in the row of the
parameter.
The Edit Analyzer dialog is displayed.
-
Perform any of the following actions to modify the default settings for the ML
analysis of the specific parameter.
If you want to... Action Disable the analysis for the specific parameter. Perform the following steps: - Click Is Enabled to
switch the analysis from being enabled or active
to become disabled or inactive.Note: If you modify the default setting to be disabled, Test Hub does not perform ML analysis of the specific parameter when tests are run subsequently.
- Click Submit.
Enable the analysis for the specific parameter, if it is disabled. Perform the following steps: - Click Is Enabled to switch the analysis from being disabled or inactive to become enabled or active.
- Click Submit.
Increase or decrease the threshold level. Perform the following steps: - Edit the value in the Decision
Threshold field and enter a value as
required.The threshold value determines the level after which the ML analysis identifies an insight into the specific parameter when the test runs.Note: You can retain the default values for the decision threshold initially until a large number of tests are run that contribute to the insights for a specific parameter. You can change the value to a higher threshold when there are a large number of records learned by the ML analyzers.
- Click Submit.
- Click Is Enabled to
switch the analysis from being enabled or active
to become disabled or inactive.
Results
You either disabled or enabled the analyzer or changed the decision threshold set for a ML analyzer.
What to do next
You can perform any of the following tasks:
- Rerun the analysis of the test with the changed threshold values without running the tests again. See Rerunning Machine learning (ML) analysis.
- Run tests if you changed the decision threshold values so that the ML analysis uses the changed values. See Test run configurations.
- View insights into the specific parameter whose decision threshold was modified after the test runs are completed. See Viewing Machine learning insights.
- View the ML recommendations for the specific parameter whose decision threshold was modified. See Viewing Machine learning analyzers.