Adaptive confidence sets for matrix completion

DOI10.3150/17-BEJ933MaRDI QIDQ1708972zbMATH OpenOpenAlexFDO

Authors Alexandra Carpentier, Olga Klopp, Matthias Löffler, Richard Nickl

Publication date 27 March 2018

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1608.04861, https://projecteuclid.org/euclid.bj/1522051214

zbMATH Keywords

matrix completionadaptivityminimax hypothesis testingconfidence setsunknown variancelow rank recovery

Mathematics Subject Classification ID

Nonparametric tolerance and confidence regions (62G15) Estimation in multivariate analysis (62H12) Minimax procedures in statistical decision theory (62C20)

Abstract: In the present paper we study the problem of existence of honest and adaptive confidence sets for matrix completion. We consider two statistical models: the trace regression model and the Bernoulli model. In the trace regression model, we show that honest confidence sets that adapt to the unknown rank of the matrix exist even when the error variance is unknown. Contrary to this, we prove that in the Bernoulli model, honest and adaptive confidence sets exist only when the error variance is known a priori. In the course of our proofs we obtain bounds for the minimax rates of certain composite hypothesis testing problems arising in low rank inference.

Recommendations

• Constructing confidence sets for the matrix completion problem
• On signal detection and confidence sets for low rank inference problems
• Confidence sets in sparse regression
• Adaptive multinomial matrix completion
• Inference and uncertainty quantification for noisy matrix completion

Cites work

• A remark on global positioning from local distances
• A sharp adaptive confidence ball for self-similar functions
• A simpler approach to matrix completion
• Accuracy assessment for high-dimensional linear regression
• Adaptive confidence sets in \(L^2\)
• Adaptive nonparametric confidence sets
• An adaptation theory for nonparametric confidence intervals
• Concentration inequalities. A nonasymptotic theory of independence
• Confidence balls in Gaussian regression.
• Confidence bands in density estimation
• Confidence sets in sparse regression
• Estimation of high-dimensional low-rank matrices
• Exact matrix completion via convex optimization
• Frequentist coverage of adaptive nonparametric Bayesian credible sets
• Honest and adaptive confidence sets in L_p
• Low-rank matrix completion using alternating minimization
• Mathematical foundations of infinite-dimensional statistical models
• Matrix completion by singular value thresholding: sharp bounds
• Matrix completion from noisy entries
• Matrix completion via max-norm constrained optimization
• Matrix estimation by universal singular value thresholding
• New concentration inequalities in product spaces
• Noisy low-rank matrix completion with general sampling distribution
• Nuclear-norm penalization and optimal rates for noisy low-rank matrix completion
• On adaptive inference and confidence bands
• On nonparametric confidence intervals
• On signal detection and confidence sets for low rank inference problems
• Recovering Low-Rank Matrices From Few Coefficients in Any Basis
• Restricted strong convexity and weighted matrix completion: optimal bounds with noise
• Sharp nonasymptotic bounds on the norm of random matrices with independent entries
• The Power of Convex Relaxation: Near-Optimal Matrix Completion
• Tight Oracle Inequalities for Low-Rank Matrix Recovery From a Minimal Number of Noisy Random Measurements
• Uncertainty quantification for matrix compressed sensing and quantum tomography problems

Cited in

(5)

• On signal detection and confidence sets for low rank inference problems
• Tight risk bound for high dimensional time series completion
• Adaptive and Implicit Regularization for Matrix Completion
• Comments on: ``High-dimensional simultaneous inference with the bootstrap
• Constructing confidence sets for the matrix completion problem

This page was built for publication: Adaptive confidence sets for matrix completion

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1708972)