Journal article
An alternate version of the conceptual predictive statistic based on a symmetrized discrepancy measure
Journal of statistical planning and inference, Vol.140(11), pp.3389-3398
2010
DOI: 10.1016/j.jspi.2010.05.002
Abstract
The conceptual predictive statistic, C
p
, is a widely used criterion for model selection in linear regression. C
p
serves as an estimator of a discrepancy, a measure that reflects the disparity between the generating model and a fitted candidate model. This discrepancy, based on scaled squared error loss, is asymmetric: an alternate measure is obtained by reversing the roles of the two models in the definition of the measure. We propose a variant of the C
p
statistic based on estimating a symmetrized version of the discrepancy targeted by C
p
. We claim that the resulting criterion provides better protection against overfitting than C
p
, since the symmetric discrepancy is more sensitive towards detecting overspecification than its asymmetric counterpart. We illustrate our claim by presenting simulation results. Finally, we demonstrate the practical utility of the new criterion by discussing a modeling application based on data collected in a cardiac rehabilitation program at University of Iowa Hospitals and Clinics.
Details
- Title: Subtitle
- An alternate version of the conceptual predictive statistic based on a symmetrized discrepancy measure
- Creators
- Joseph E Cavanaugh - Department of Biostatistics, C22 GH, 200 Hawkins Drive, The University of Iowa, Iowa City, IA 52242, USAAndrew A Neath - Department of Mathematics and Statistics, Southern Illinois University Edwardsville, USASimon L Davies - Pfizer Global Pharmaceuticals, Inc., USA
- Resource Type
- Journal article
- Publication Details
- Journal of statistical planning and inference, Vol.140(11), pp.3389-3398
- DOI
- 10.1016/j.jspi.2010.05.002
- ISSN
- 0378-3758
- eISSN
- 1873-1171
- Publisher
- Elsevier B.V
- Language
- English
- Date published
- 2010
- Academic Unit
- Statistics and Actuarial Science; Biostatistics; Injury Prevention Research Center
- Record Identifier
- 9984214688002771
Metrics
27 Record Views