Error estimation and model selection

Correct model selection requires careful data handling and attention to iterative or cyclic design / deployment procedures, as are the norm in everyday business applications. Performance estimates must be revised in view of data (re-)sampling and model selection procedures.
Check all of our work