The marginalized likelihood ratio test for detecting abrupt changes

Author

Gustafsson, Fredrik

Author_Institution

Dept. of Electr. Eng., Linkoping Univ., Sweden

Volume

41

Issue

1

fYear

1996

fDate

1/1/1996 12:00:00 AM

Firstpage

66

Lastpage

78

Abstract

The generalized likelihood ratio (GLR) test is a widely used method for detecting abrupt changes in linear systems and signals. In this paper the marginalized likelihood ratio (MLR) test is introduced for eliminating three shortcomings of GLR while preserving its applicability and generality. First, the need for a user-chosen threshold is eliminated in MLR. Second, the noise levels need not be known exactly and may even change over time, which means that MLR is robust. Finally, a very efficient exact implementation with linear in time complexity for batch-wise data processing is developed. This should be compared to the quadratic in time complexity of the exact GLR

Keywords

computational complexity; estimation theory; filtering theory; linear systems; parameter estimation; state-space methods; abrupt change detection; batch-wise data processing; generalized likelihood ratio; linear systems; marginalized likelihood ratio test; time complexity; Data processing; Estimation theory; Filters; Linear regression; Linear systems; Noise level; Noise robustness; Stochastic resonance; Stochastic systems; Testing;

fLanguage

English

Journal_Title

Automatic Control, IEEE Transactions on

Publisher

ieee

ISSN

0018-9286

Type

jour

DOI

10.1109/9.481608

Filename

481608