Parameter estimation and order determination in the low-rank linear statistical model

Document Type

Conference Proceeding

Date of Original Version



We consider the problem of estimating the min-norm solution to a low-rank, linear statistical model. We calculate the statistics of the solution as a function of the statistical characterization of the matrix containing observation noise. We also present a new method for estimating the rank of the underlying noise-free matrix.

Publication Title, e.g., Journal

IEEE International Symposium on Information Theory - Proceedings

This document is currently not available here.