Title

On the condition number of Gaussian sample-covariance matrices

Authors

Alexander Kostinski, Michigan Technological UniversityFollow
A. C. Koivunen

Document Type

Article

Publication Date

1-2000

Abstract

The authors examine the reasons behind the fact that the Gaussian autocorrelation-function model, widely used in remote sensing, yields a particularly ill-conditioned sample-covariance matrix in the case of many strongly correlated samples. The authors explore the question numerically and relate the magnitude of the matrix-condition number to the nonnegativity requirement satisfied by all correlation functions. They show that the condition number exhibits explosive growth near the boundary of the allowed-parameter space, simple numerical recipes are suggested in order to avoid this instability.

Publisher's Statement

© 2000 IEEE. Publisher’s version of record: https://dx.doi.org/10.1109/36.823928

Publication Title

IEEE Transactions on Geoscience and Remote Sensing

Recommended Citation

Kostinski, A., & Koivunen, A. C. (2000). On the condition number of Gaussian sample-covariance matrices. IEEE Transactions on Geoscience and Remote Sensing, 38(1), 329-332. http://doi.org/10.1109/36.823928
Retrieved from: https://digitalcommons.mtu.edu/physics-fp/241