Mutual information for low-rank even-order symmetric tensor estimation

DOI10.1093/IMAIAI/IAAA022MaRDI QIDQ5033273zbMATH OpenOpenAlexFDO

Authors Clément Luneau, Jean Barbier, Nicolas Macris

Publication date 22 February 2022

Published in Information and Inference: A Journal of the IMA (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1904.04565

zbMATH Keywords

adaptive interpolation methodconcentrationmutual informationreplica formulatensor decomposition

Mathematics Subject Classification ID

Statistical aspects of information-theoretic topics (62B10) Estimation in multivariate analysis (62H12) Information theory (general) (94A15)

Abstract: We consider a statistical model for finite-rank symmetric tensor factorization and prove a single-letter variational expression for its asymptotic mutual information when the tensor is of even order. The proof applies the adaptive interpolation method originally invented for rank-one factorization. Here we show how to extend the adaptive interpolation to finite-rank and even-order tensors. This requires new nontrivial ideas with respect to the current analysis in the literature. We also underline where the proof falls short when dealing with odd-order tensors.

Recommendations

• Statistical inference of finite-rank tensors
• Fundamental limits of symmetric low-rank matrix estimation
• Statistical limits of spiked tensor models
• Statistical mechanics of low-rank tensor decomposition
• Statistical thresholds for tensor PCA

Cited in

(11)

• Hamilton-Jacobi equations for nonsymmetric matrix inference
• Sparse high-dimensional linear regression. Estimating squared error and a phase transition
• Hamilton-Jacobi equations for inference of matrix tensor products
• Mutual information for the sparse stochastic block model
• Free energy of multi-layer generalized linear models
• Statistical inference of finite-rank tensors
• Estimating rank-one matrices with mismatched prior and noise: universality and large deviations
• Central limit theorem for the overlaps in mean-field spin glasses on the Nishimori line
• A multiscale cavity method for sublinear-rank symmetric matrix factorization
• Differentiability and overlap concentration in optimal Bayesian inference
• Pseudo-maximum likelihood theory for high-dimensional rank one inference

This page was built for publication: Mutual information for low-rank even-order symmetric tensor estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5033273)