Abstract
The optimal transport (OT) problem has gained significant traction in modernmachine learning for its ability to: (1) provide versatile metrics, such asWasserstein distances and their variants, and (2) determine optimal couplingsbetween probability measures. To reduce the computational complexity of OTsolvers, methods like entropic regularization and sliced optimal transport havebeen proposed. The sliced OT framework improves efficiency by comparingone-dimensional projections (slices) of high-dimensional distributions.However, despite their computational efficiency, sliced-Wasserstein approacheslack a transportation plan between the input measures, limiting their use inscenarios requiring explicit coupling. In this paper, we address two keyquestions: Can a transportation plan be constructed between two probabilitymeasures using the sliced transport framework? If so, can this plan be used todefine a metric between the measures? We propose a "lifting" operation toextend one-dimensional optimal transport plans back to the original space ofthe measures. By computing the expectation of these lifted plans, we derive anew transportation plan, termed expected sliced transport (EST) plans. We provethat using the EST plan to weight the sum of the individual Euclidean costs formoving from one point to another results in a valid metric between the inputdiscrete probability measures. We demonstrate the connection between ourapproach and the recently proposed min-SWGG, along with illustrative numericalexamples that support our theoretical findings.