I also liked a lot this ICML paper [1], which is more theoretically principled than the two you reference (and weirdly they do not cite it). [1] http://www.eec

I also liked a lot this ICML paper [1], which is more theoretically principled than the two you reference (and weirdly they do not cite it). [1] http://www.eecs.tufts.edu/~dsculley/papers/round-model-icml.pdf http://www.eecs.tufts.edu/~dsculley/papers/round-model-icml....