On the condition number of the total least squares problem with linear equality constraint
share
This paper is devoted to the condition number of the total least squares problem with linear equality constraint (TLSE). With novel techniques, closed formulae for the normwise, mixed and componentwise condition numbers of the TLSE problem are derived. Compact expressions and upper bounds for these condition numbers are also given to avoid the costly Kronecker product-based operations. Explicit condition number expressions and perturbation bound for the TLS problem can be recovered from our estimates. For random TLSE problems, numerical experiments illustrate the sharpness of the estimates based on normwise condition numbers, while for sparse and badly scaled matrices, the estimates based on mixed and componentwise condition numbers are much tighter.
READ FULL TEXT