A Meta-Analysis of Overfitting in Machine Learning

Rebecca Roelofs, Vaishaal Shankar, Benjamin Recht, Sara Fridovich-Keil, Moritz Hardt, John Miller, Ludwig Schmidt

Neural Information Processing Systems 

In each competition, numerous practitioners repeatedly evaluated their progress against a holdout set that forms the basis of a public ranking availablethroughout the competition. Performance on a separate test set used only oncedetermined the final ranking.

Similar Docs  Excel Report  more

TitleSimilaritySource
None found