A bootstrap method for structure detection of NARMAX models (Q4474753)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 2078892
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | A bootstrap method for structure detection of NARMAX models |
scientific article; zbMATH DE number 2078892 |
Statements
A bootstrap method for structure detection of NARMAX models (English)
0 references
12 July 2004
0 references
nonlinear NARMAX models
0 references
structure detection
0 references
bootstrap technique
0 references
robustness
0 references
extended least-squares parameter estimation
0 references
Monte Carlo simulation
0 references
0 references
0 references
0 references
0.8665232
0 references
0.8532998
0 references
0.8420682
0 references
0.84060884
0 references
0.8372771
0 references
0.8364496
0 references
0.83612555
0 references
0.83501256
0 references
The paper presents a bootstrap technique combined with an extended least-squares parameter estimation algorithm as a means for structure detection of nonlinear NARMAX models. The performance of the proposed bootstrap structure detection algorithm is examined by extensive Monte Carlo simulations. The presented simulation examples demonstrate that in the case of uniformly distributed, white and zero-mean inputs, and under Gaussian, zero-mean white output noise, the proposed algorithm is a robust structure detection technique, yielding parsimonious NARMAX models. The method is particularly efficient for highly over-parameterized models and under poor a priori knowledge about the true system.
0 references