The least squares: output error sensitivity and the constrained logarithmic algorithm

Er-Wei Bai; Yinyu Ye

doi:10.1109/ACC.1998.703277

Back

Conference proceeding

The least squares: output error sensitivity and the constrained logarithmic algorithm

Er-Wei Bai and Yinyu Ye

Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207), Vol.6, pp.3570-3574 vol.6

1998

DOI: 10.1109/ACC.1998.703277

View Online

Abstract

It is shown in the paper that while robust in terms of the average output error, the least squares estimate is sensitive to outliers with respect to the maximum output error. In fact, the worst-case output error of the least squares can go unbounded. The constrained logarithmic least squares for system identification is applied to solve the problem.

Cities and towns

Computer errors

Engineering management

Least squares approximation

Least squares methods

Noise measurement

Q measurement

Robustness

Size measurement

Transfer functions

Details

Title: Subtitle: The least squares: output error sensitivity and the constrained logarithmic algorithm
Creators: Er-Wei Bai - Dept. of Electr. & Comput. Eng., Iowa Univ., Iowa City, IA, USA
Yinyu Ye
Resource Type: Conference proceeding
Publication Details: Proceedings of the 1998 American Control Conference. ACC (IEEE Cat. No.98CH36207), Vol.6, pp.3570-3574 vol.6
Publisher: IEEE
DOI: 10.1109/ACC.1998.703277
ISSN: 0743-1619
eISSN: 2378-5861
Language: English
Date published: 1998
Academic Unit: Electrical and Computer Engineering
Record Identifier: 9984197518002771

Metrics

16 Record Views