Modified AIC and MDL model selection criteria for short data records
- Resource Type
- Periodical
- Authors
- De Ridder, F.; Pintelon, R.; Schoukens, J.; Gillikin, D.P.
- Source
- IEEE Transactions on Instrumentation and Measurement IEEE Trans. Instrum. Meas. Instrumentation and Measurement, IEEE Transactions on. 54(1):144-150 Feb, 2005
- Subject
- Power, Energy and Industry Applications
Components, Circuits, Devices and Systems
Autoregressive processes
Cost function
Parameter estimation
Noise measurement
Signal processing
Gaussian noise
Length measurement
Time measurement
Noise reduction
Nonlinear dynamical systems
Akaike information criterion (AIC)
finite sample
minimum description length (MDL)
model selection
- Language
- ISSN
- 0018-9456
1557-9662
The classical model selection rules such as Akaike information criterion (AIC) and minimum description length (MDL) have been derived assuming that the number of samples (measurements) is much larger than the number of estimated model parameters. For short data records, AIC and MDL have the tendency to select overly complex models. This paper proposes modified AIC and MDL rules with improved finite sample behavior. They are useful in those measurement applications where gathering a sample is very time consuming and/or expensive.