Modified AIC rule for model selection in combination with prior estimated noise models (Q1614314): Difference between revisions
From MaRDI portal
Set profile property. |
ReferenceBot (talk | contribs) Changed an Item |
||
Property / cites work | |||
Property / cites work: A new look at the statistical model identification / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Modeling by shortest data description / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Q3657269 / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Frequency-domain system identification using non-parametric noise models estimated from a small number of data sets / rank | |||
Normal rank | |||
Property / cites work | |||
Property / cites work: Frequency domain system identification with missing data / rank | |||
Normal rank |
Latest revision as of 16:42, 4 June 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Modified AIC rule for model selection in combination with prior estimated noise models |
scientific article |
Statements
Modified AIC rule for model selection in combination with prior estimated noise models (English)
0 references
5 September 2002
0 references
The authors discuss the use of the AIC model selection rule in combination with prior estimated noise models when model errors are present. When a prior estimated noise model is used as a weighting in the cost function, it is shown that a multiplicative model complexity penalty term is needed. The connexion of this result with the classical prediction error framework is examined.
0 references
AIC
0 references
model selection
0 references
model errors
0 references
0 references