On the Wasserstein median of probability measures

Abstract

The primary choice to summarize a finite collection of random objects is by using measures of central tendency, such as mean and median. In the field of optimal transport, the Wasserstein barycenter corresponds to the Fréchet or geometric mean of a set of probability measures, which is defined as a minimizer of the sum of squared distances to each element in a given set with respect to the Wasserstein distance of order 2. We introduce the Wasserstein median as a robust alternative to the Wasserstein barycenter. The Wasserstein median corresponds to the Fréchet median under the 2-Wasserstein metric. The existence and consistency of the Wasserstein median are first established, along with its robustness property. In addition, we present a general computational pipeline that employs any recognized algorithms for the Wasserstein barycenter in an iterative fashion and demonstrate its convergence. The utility of the Wasserstein median as a robust measure of central tendency is demonstrated using real and simulated data.

40 pages, 16 figures

References in corpus (8)

• Bayesian Inference with Optimal Maps
• Geometric median and robust estimation in Banach spaces
• Approximate Bayesian computation with the Wasserstein distance
• Joint Distribution Optimal Transportation for Domain Adaptation
• Scalable Bayes via Barycenter in Wasserstein Space
• Parameter Estimation and Model-Based Clustering with Spherical Normal Distribution on the Unit Hypersphere
• Continuous Wasserstein-2 Barycenter Estimation without Minimax Optimization
• Averaging on the Bures-Wasserstein manifold: dimension-free convergence of gradient descent

Cited by in corpus (2)

• Automated regime classification in multidimensional time series data using sliced Wasserstein k-means clustering
• Geometric medians on product manifolds