When should data scientists try a new technique?
MIT researchers created a new measure, called the c-value, that helps statisticians and data scientists choose between estimation techniques based on the chance that a new method is more accurate for a specific dataset and application.