Abstract | ||
---|---|---|
Biased estimation has the advantage of reducing the Mean Squared Error (MSE) of an estimator. The question of interest is how biased estimation affects model selection. In this paper, we introduce biased estimation to a range of model selection criteria. Specifically, we analyze the performance of the Minimum Description Length (MDL) criterion based on biased and unbiased estimation and compare it against modern model selection criteria such as Kay's conditional model order estimator (CME), the Bootstrap and the more recently proposed Hook-and-Loop resampling based model selection. The advantages and limitations of the considered techniques are discussed. The results indicate that, in some cases, biased estimators can slightly improve the selection of the correct model. We also give an example for which the CME with an unbiased estimator fails, but could regain its power when a biased estimator is used. |
Year | DOI | Venue |
---|---|---|
2009 | 10.1109/ICASSP.2009.4960370 | ICASSP |
Keywords | DocType | ISSN |
mean squared error,model order selection,unbiased estimation,minimum description length,modern model selection criterion,conditional model order estimator,correct model,unbiased estimator,model selection criterion,model selection,data mining,estimation theory,mean square error,estimation,covariance matrix,context modeling,sampling methods,polynomials,bootstrap,vectors,kelvin,signal to noise ratio,lenses,mathematical model | Conference | 1520-6149 |
Citations | PageRank | References |
0 | 0.34 | 1 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Weaam Alkhaldi | 1 | 0 | 1.35 |
D. Robert Iskander | 2 | 104 | 23.65 |
Abdelhak M. Zoubir | 3 | 1036 | 148.03 |