Trace Inequalities with Applications to Orthogonal Regression and Matrix Nearness Problems
Ian Coope and Peter Renaud
View Report [PDF - 104 KB]
Abstract
Matrix trace inequalities are finding increased use in many areas such as analysis, where they can be used to generalise several well known classical inequalities, and computational statistics, where they can be applied, for example, to data fitting problems. In this paper we give simple proofs of two useful matrix trace inequalities and provide applications to orthogonal regression and matrix nearness problems.
Back to Research Reports