Predictive Test Selection
Simulate
Model performance curves
You can use these curves to choose an optimization target for your subset test runs. Learn more
Confidence curve
A confidence curve tells you the percentage of failing test runs you should expect to catch based on how much time has been spent running tests. How we generate this curve
The Launchable model line shows the expected performance of this workspace’s machine learning model.
The maximum value of the x-axis (Test execution time) equals the duration of the longest test session used to evaluate this model.
Your model’s curves will evolve over time based on new builds and test results.