- Title
- On the worst-case divergence of the least-squares algorithm
- Creator
- Ackay, Huseyin; Ninness, Brett
- Relation
- Systems & Control Letters Vol. 33, Issue 1, p. 19-24
- Publisher Link
- http://dx.doi.org/10.1016/S0167-6911(97)00093-5
- Publisher
- Elsevier Science
- Resource Type
- journal article
- Date
- 1998
- Description
- In this paper, we provide a H ∞ norm lower bound on the worst-case identification error of least-squares estimation when using FIR model structures. This bound increases as a logarithmic function of model complexity and is valid for a wide class of inputs characterized as being quasi-stationary with covariance function falling off sufficiently quickly.
- Subject
- least-squares; identification in H ∞; time-domain data; divergance
- Identifier
- http://hdl.handle.net/1959.13/30303
- Identifier
- uon:2668
- Identifier
- ISSN:0167-6911
- Language
- eng
- Full Text
- Reviewed
- Hits: 5806
- Visitors: 6190
- Downloads: 449
| Thumbnail | File | Description | Size | Format | |||
|---|---|---|---|---|---|---|---|
| View Details Download | ATTACHMENT01 | Author final version | 189 KB | Adobe Acrobat PDF | View Details Download |