Groups
Category
An f-divergence measures how different two probability distributions P and Q are by averaging a convex function f of the density ratio p(x)/q(x) under Q.
Transfer learning theory studies when and why a model trained on a source distribution will work on a different target distribution.