Poisson Matrix Recovery and Completion

DOI10.1109/TSP.2015.2500192MaRDI QIDQ4618269zbMATH OpenOpenAlexFDO

Authors Yang Cao, Yao Xie

Publication date 7 February 2019

Published in IEEE Transactions on Signal Processing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1504.05229

Mathematics Subject Classification ID

Signal theory (characterization, reconstruction, filtering, etc.) (94A12)

Abstract: We extend the theory of low-rank matrix recovery and completion to the case when Poisson observations for a linear combination or a subset of the entries of a matrix are available, which arises in various applications with count data. We consider the usual matrix recovery formulation through maximum likelihood with proper constraints on the matrix $M$ of size $d_1$-by-$d_2$, and establish theoretical upper and lower bounds on the recovery error. Our bounds for matrix completion are nearly optimal up to a factor on the order of $mathcalO(log(d_1{d}_2))$. These bounds are obtained by combing techniques for compressed sensing for sparse vectors with Poisson noise and for analyzing low-rank matrices, as well as adapting the arguments used for one-bit matrix completion cite{davenport20121} (although these two problems are different in nature) and the adaptation requires new techniques exploiting properties of the Poisson likelihood function and tackling the difficulties posed by the locally sub-Gaussian characteristic of the Poisson distribution. Our results highlight a few important distinctions of the Poisson case compared to the prior work including having to impose a minimum signal-to-noise requirement on each observed entry and a gap in the upper and lower bounds. We also develop a set of efficient iterative algorithms and demonstrate their good performance on synthetic examples and real data.

Recommendations

• Low-rank matrix completion and denoising under Poisson noise
• The P₀-matrix completion problem
• The P₀^+-matrix completion problem
• Reconstruction of multidimensional distribution with Poisson components
• Log-sum penalized Poisson loss minimization for matrix recovery
• Matrix completion problems
• Structured matrix estimation and completion
• scientific article; zbMATH DE number 5002128
• Matrix Completion Theorems

Cited in

(22)

• Main effects and interactions in mixed and incomplete data frames
• Reconstruction of multidimensional distribution with Poisson components
• Cross: efficient low-rank tensor completion
• Low-rank matrix denoising for count data using unbiased Kullback-Leibler risk estimation
• An optimal statistical and computational framework for generalized tensor estimation
• Low-rank model with covariates for count data with missing values
• Generalized SURE for optimal shrinkage of singular values in low-rank matrix denoising
• Learning Markov models via low-rank optimization
• Noisy matrix completion: understanding statistical guarantees for convex relaxation via nonconvex optimization
• Low-rank matrix completion and denoising under Poisson noise
• Biwhitening Reveals the Rank of a Count Matrix
• A Zero-Inflated Logistic Normal Multinomial Model for Extracting Microbial Compositions
• Exponential family tensor completion with auxiliary information
• Empirical Bayes Poisson matrix completion
• Group projected subspace pursuit for block sparse signal reconstruction: convergence analysis and applications
• A composite optimization algorithm for Poisson tensor completions without nonnegative constraints based on the generalized Gauss-Newton method
• High-dimensional large-scale mixed-type data imputation under missing at random
• Low-rank matrix completion with Poisson observations via nuclear norm and total variation constraints
• Poisson tensor completion via nonconvex regularization and nonlocal self-similarity for multi-dimensional image recovery
• Unbiased test error estimation in the Poisson means problem via coupled bootstrap techniques
• Statistical inference in tensor completion: optimal uncertainty quantification and statistical-to-computational gaps
• Optimal rates for smooth nonlinear matrix completion via low-rank approximation

This page was built for publication: Poisson Matrix Recovery and Completion

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4618269)