no code implementations • 16 Apr 2020 • Michael Gleicher, Aditya Barve, Xinyi Yu, Florian Heimerl
The approach focuses on allowing the user to identify interesting subsets of training and testing instances and comparing performance of the classifiers on these subsets.