Title :
Data validation in the presence of imprecisely known correlations
Author :
Hanebeck, U.D. ; Horn, J.
Author_Institution :
Inst. of Computer Design and Fault Tolerance, Universität Karlsruhe, 76128 Karlsruhe, Germany
Abstract :
This paper derives fundamental results for data validation in the presence of imprecisely known correlations. Given a constraint on the maximum absolute correlation of a given estimate and measurement data, a tight upper bound for the joint covariance matrix is derived, which finally yields a modified Mahalanobis distance. The special cases of one-dimensional and two-dimensional random variables are discussed.
Keywords :
Correlation; Covariance matrices; Ellipsoids; Joints; Measurement uncertainty; Random variables; Upper bound; Covariance Bounds; Data Validation; Imprecisely Known Correlations; Mahalanobis Distance; Stochastic Uncertainties;
Conference_Titel :
European Control Conference (ECC), 2003
Conference_Location :
Cambridge, UK
Print_ISBN :
978-3-9524173-7-9
