Using machine learning to determine which automated tests are most likely to fail based on code changes and selecting only those tests to run. Predictive test selection is a method of Test Impact Analysis.
If your tests take a long time to run, you can use predictive test selection to determine which tests are most relevant to your changes and only run those tests. This can dramatically reduce the time required for a test run.
Both Google and Facebook have used predictive test selection to great effect to reduce massive test suites to the most relevant tests. Launchable’s adaptive subsets productize predictive test selection so that you can use it with almost any test suite, without having to understand the intricacies of machine learning or maintaining costly infrastructure.
Predictive test selection can be seen as an alternative to test parallelization and manual test selection, although it can be combined with either technique.