A variety of important machine learning applications require predictions on test data with different characteristics than the data on which a model was trained and validated. In particular, test data may have a different relative frequency of positives and negatives (i.e., class distribution) and/or different mislabeling costs of false positive...
Software entropy impacts the overall quality of software systems. High entropy hinders developers from understanding the purpose of a piece of code and can cause developers to make sub-optimal changes and introduce bugs. Researchers have used entropy scores to measure the naturalness of code. However, thus far, no one has...