Abstract

In this paper we consider the low-rank matrix completion problem withspecific application to forecasting in time series analysis. Briefly, thelow-rank matrix completion problem is the problem of imputing missing values ofa matrix under a rank constraint. We consider a matrix completion problem forHankel matrices and a convex relaxation based on the nuclear norm. Based on newtheoretical results and a number of numerical and real examples, we investigatethe cases when the proposed approach can work. Our results highlight theimportance of choosing a proper weighting scheme for the known observations.

 

Quick Read (beta)