TS_benchmarking_subset
A project to find an efficient subset of datasets that would give an approximation of some benchmark calculated on a large number of datasets. The minimum subset of datasets should give approximately the same ranks (in terms of, for example, correlation) for model predictions.
