NYou are working on a supervised learning model and the current version is not meeting the performance bar. Several candidate algorithms and preprocessing choices are on the table, and you need a principled way to improve results without overfitting.
How would you optimize a machine learning model for performance?
