of overparametrization in non-linear system identification and neural networks (Q4784556)
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 1844125
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | of overparametrization in non-linear system identification and neural networks |
scientific article; zbMATH DE number 1844125 |
Statements
of overparametrization in non-linear system identification and neural networks (English)
0 references
16 October 2003
0 references
NARMAX models
0 references
overparametrization
0 references
nonlinear system identification
0 references
neural networks
0 references
overfitting
0 references
qualitative validation
0 references
0 references
0 references
0.87726766
0 references
0.8745372
0 references
0.8728664
0 references
0.87126034
0 references
0.8693712
0 references
0.86377096
0 references
0.86158204
0 references
The effects of overparametrization in nonlinear system identification and neural networks are often presented as overfitting to a particular data set. This paper studies and compares the effects of several forms of overparametrization on the model dynamics. For that, qualitative validation methods are used. The dynamic properties of identified model or the brained network are graphically displayed and compared with those of the real system. It is shown that overparametrization in the number of time lags or input nodes has the worst effect on the dynamic properties of the resulting model.
0 references