Statistical inference for Bures–Wasserstein barycenters

In this work we introduce the concept of Bures-Wasserstein barycenter $Q_*$, that is essentially a Frechet mean of some distribution $\mathbb{P}$ supported on a subspace of positive semi-definite Hermitian operators $\mathbb{H}_{+}(d)$. We allow a barycenter to be restricted to some affine subspace of $\mathbb{H}_{+}(d)$ and provide conditions ensuring its existence and uniqueness. We also investigate convergence and concentration properties of an empirical counterpart of $Q_*$ in both Frobenius norm and Bures-Wasserstein distance, and explain, how obtained results are connected to optimal transportation theory and can be applied to statistical inference in quantum mechanics.

[1]  Thibaut Le Gouic,et al.  Fast convergence of empirical barycenters in Alexandrov spaces and the Wasserstein space , 2019, Journal of the European Mathematical Society.

[2]  E. Barrio,et al.  A statistical analysis of a deformation model with Wasserstein barycenters : estimation procedure and goodness of fit test , 2015, 1508.06465.

[3]  I. Olkin,et al.  The distance between two random vectors with given dispersion matrices , 1982 .

[4]  Aasa Feragen,et al.  Learning from uncertain curves: The 2-Wasserstein metric for Gaussian processes , 2017, NIPS.

[5]  U. Fano.  Description of States in Quantum Mechanics by Density Matrix and Operator Techniques , 1957 .

[6]  Jérémie Bigot,et al.  Consistent estimation of a population barycenter in the Wasserstein space , 2013 .

[7]  Klaus-Robert Müller,et al.  Wasserstein Training of Restricted Boltzmann Machines , 2016, NIPS.

[8]  Y. Brenier.  Polar Factorization and Monotone Rearrangement of Vector-Valued Functions , 1991 .

[9]  J. A. Cuesta-Albertos,et al.  Wide consensus aggregation in the Wasserstein space. Application to location-scatter families , 2018, Bernoulli.

[10]  Gabriel Peyré,et al.  Fast Optimal Transport Averaging of Neuroimaging Data , 2015, IPMI.

[11]  Marco Cuturi,et al.  Sinkhorn Distances: Lightspeed Computation of Optimal Transport , 2013, NIPS.

[12]  V. Koltchinskii.  Von Neumann Entropy Penalization and Low Rank Matrix Estimation , 2010, 1009.2439.

[13]  Charles J. Goodnight,et al.  A Bootstrap Comparison of Genetic Covariance Matrices , 1997 .

[14]  Jean-Michel Loubes,et al.  Central limit theorem and bootstrap procedure for Wasserstein's variations with an application to structural relationships between distributions , 2016, J. Multivar. Anal..

[15]  Darina Dvinskikh,et al.  Decentralize and Randomize: Faster Algorithm for Wasserstein Barycenters , 2018, NeurIPS.

[16]  C. Goodnight,et al.  Empirical Comparison of G Matrix Test Statistics: Finding Biologically Relevant Change , 2009, Evolution; international journal of organic evolution.

[17]  L. Ambrosio,et al.  A User’s Guide to Optimal Transport , 2013 .

[18]  P. Marian,et al.  Bures distance as a measure of entanglement for symmetric two-mode Gaussian states , 2007, 0705.1138.

[19]  Nicolas Courty,et al.  Wasserstein discriminant analysis , 2016, Machine Learning.

[20]  Jerzy Dajka,et al.  Distance between quantum states in the presence of initial qubit-environment correlations: A comparative study , 2011, 1107.1732.

[21]  Asuka Takatsu.  Wasserstein geometry of Gaussian measures , 2011 .

[22]  Marco Cuturi,et al.  Generalizing Point Embeddings using the Wasserstein Space of Elliptical Distributions , 2018, NeurIPS.

[23]  Sham M. Kakade,et al.  A tail inequality for quadratic forms of subgaussian random vectors , 2011, ArXiv.

[24]  R. Tatusko.  Cooperation in climate research : an evaluation of the activities conducted under the US-USSR agreement for environmental protection since 1974 , 1990 .

[25]  R. Jozsa.  Fidelity for Mixed Quantum States , 1994 .

[26]  Alexey Kroshnin.  Fréchet barycenters in the Monge-Kantorovich spaces , 2017, 1702.05740.

[27]  Thibaut Le Gouic,et al.  Existence and consistency of Wasserstein barycenters , 2015, Probability Theory and Related Fields.

[28]  Axel Munk,et al.  Limit laws of the empirical Wasserstein distance: Gaussian distributions , 2015, J. Multivar. Anal..

[29]  E. Barrio,et al.  Central limit theorems for empirical transportation cost in general dimension , 2017, The Annals of Probability.

[30]  Martial Agueh,et al.  Vers un théorème de la limite centrale dans l'espace de Wasserstein ? , 2017 .

[31]  Thibaut Le Gouic,et al.  On the rate of convergence of empirical barycentres in metric spaces: curvature, convexity and extendible geodesics , 2019 .

[32]  Nicolas Courty,et al.  Optimal Transport for Domain Adaptation , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[33]  Guillaume Carlier,et al.  Barycenters in the Wasserstein Space , 2011, SIAM J. Math. Anal..

[34]  Rachid Deriche,et al.  Unsupervised white matter fiber clustering and tract probability map generation: Applications of a Gaussian process framework for white matter fibers , 2010, NeuroImage.

[35]  Oscar Gonzalez,et al.  Absolute versus Relative Entropy Parameter Estimation in a Coarse-Grain Model of DNA , 2017, Multiscale Model. Simul..

[36]  V. Spokoiny,et al.  Construction of Non-asymptotic Confidence Sets in 2-Wasserstein Space , 2017, 1703.03658.

[37]  Darina Dvinskikh,et al.  On the Complexity of Approximating Wasserstein Barycenters , 2019, ICML.

[38]  Juan Antonio Cuesta-Albertos,et al.  Robust clustering tools based on optimal transportation , 2016, Statistics and Computing.

[39]  S. R. Jammalamadaka,et al.  Empirical Processes in M-Estimation , 2001 .

[40]  R. Bhatia,et al.  On the Bures–Wasserstein distance between positive definite matrices , 2017, Expositiones Mathematicae.