We describe a randomized Krylov-subspace method for estimating the spectral conditionnumber of a real matrix A or indicating that it is numerically rank deficient. The main difficulty in estimating the condition numb...
详细信息
We describe a randomized Krylov-subspace method for estimating the spectral conditionnumber of a real matrix A or indicating that it is numerically rank deficient. The main difficulty in estimating the conditionnumber is the estimation of the smallest singular value sigma min of A. Our method estimates this value by solving a consistent linear least squares problem with a known solution using a specific Krylov-subspace method called LSQR. In this method, the forward error tends to concentrate in the direction of a right singular vector corresponding to sigma min. Extensive experiments show that the method is able to estimate well the conditionnumber of a wide array of matrices. It can sometimes estimate the conditionnumber when running dense singular value decomposition would be impractical due to the computational cost or the memory requirements. The method uses very little memory (it inherits this property from LSQR), and it works equally well on square and rectangular matrices.
暂无评论