NEAR: A Training-Free Pre-Estimator of Machine Learning Model Performance.

Raphael T. Husistein, Markus Reiher, Marco Eckhoff

Browse the full ICLR paper archive.