Evaluating ML classifiers using relative error instead of absolute accuracystephantul.github.io1 point·stephantul··0 commentsOpen articleSaveView on HN