ParentFull threadSpaceManNabs·is there a ranking of the methods that actually work on benchmark datasets? Hybrid, "ML" or old stats? I remember eamonnkeogh doing this on r/ML a few years ago.View on HN