Principal Geodesic Analysis for Probability Measures under the Optimal Transport Metric

Abstract

Given a family of probability measures in P(X), the space of probability measures on a Hilbert space X, our goal in this paper is to highlight one ore more curves in P(X) that summarize efficiently that family. We propose to study this problem under the optimal transport (Wasserstein) geometry, using curves that are restricted to be geodesic segments under that metric. We show that concepts that play a key role in Euclidean PCA, such as data centering or orthogonality of principal directions, find a natural equivalent in the optimal transport geometry, using Wasserstein means and differential geometry. The implementation of these ideas is, however, computationally challenging. To achieve scalable algorithms that can handle thousands of measures, we propose to use a relaxed definition for geodesics and regularized optimal transport distances. The interest of our approach is demonstrated on images seen either as shapes or color histograms.

9 pages, 8 figures. To appear in Advances in Neural Information Processing Systems (NIPS) 2015

References in corpus (1)

• Fast Computation of Wasserstein Barycenters

Cited by in corpus (9)

• Robust Wasserstein Profile Inference and Applications to Machine Learning
• Wasserstein Discriminant Analysis
• Wasserstein Distributionally Robust Optimization: Theory and Applications in Machine Learning
• Optimal Transport losses and Sinkhorn algorithm with general convex regularization
• LCS Graph Kernel Based on Wasserstein Distance in Longest Common Subsequence Metric Space
• Randomised Wasserstein Barycenter Computation: Resampling with Statistical Guarantees
• Wasserstein Embedding for Graph Learning
• Matching Distributions via Optimal Transport for Semi-Supervised Learning
• Regression analysis of distributional data through Multi-Marginal Optimal transport