Metric learning by directly minimizing the k-NN training error
More Info
expand_more
expand_more