Minimax statistical learning with Wasserstein distances

Recently, however, an alternative viewpoint has emerged, inspired by ideas from robust statistics and robust stochastic optimization.