Differentiable Sorting using Optimal Transport:The Sinkhorn CDF and Quantile Operator
share
Sorting an array is a fundamental routine in machine learning, one that is used to compute rank-based statistics, cumulative distribution functions (CDFs), quantiles, or to select closest neighbors and labels. The sorting function is however piece-wise constant (the sorting permutation of a vector does not change if the entries of that vector are infinitesimally perturbed) and therefore has no gradient information to back-propagate. We propose a framework to sort elements that is algorithmically differentiable. We leverage the fact that sorting can be seen as a particular instance of the optimal transport (OT) problem on R, from input values to a predefined array of sorted values (e.g. 1,2,...,n if the input array has n elements). Building upon this link , we propose generalized CDFs and quantile operators by varying the size and weights of the target presorted array. Because this amounts to using the so-called Kantorovich formulation of OT, we call these quantities K-sorts, K-CDFs and K-quantiles. We recover differentiable algorithms by adding to the OT problem an entropic regularization, and approximate it using a few Sinkhorn iterations. We call these operators S-sorts, S-CDFs and S-quantiles, and use them in various learning settings: we benchmark them against the recently proposed neuralsort [Grover et al. 2019], propose applications to quantile regression and introduce differentiable formulations of the top-k accuracy that deliver state-of-the art performance.
READ FULL TEXT
