Week 4

Model training

Two models are trained with the full dataset:

• testfailure predicts whether a commit will fail any unit test. It only considers commit features.
• testselect predicts which unit tests the commit will fail. It only considers unit tests features.

These two models are based on bugbug, but they have one main limitation. Commit and unit test features are considered independently. The better way to solve this problem is to consider these two kinds of features together to predict whether a (commit, test) pair will pass or fail.