Asymptotic bootstrap corrections of AIC for linear regression models Export

@article{Seghouane2010, abstract = {The Akaike information criterion, AIC, and its corrected version, AIC c are two methods for selecting normal linear regression models. Both criteria were designed as estimators of the expected Kullback–Leibler information between the model generating the data and the approximating candidate model. In this paper, two new corrected variants of AIC are derived for the purpose of small sample linear regression model selection. The proposed variants of AIC are based on asymptotic approximation of bootstrap type estimates of Kullback–Leibler information. These new variants are of particular interest when the use of bootstrap is not really justified in terms of the required calculations. As its the case for AIC c , these new variants are asymptotically equivalent to AIC. Simulation results which illustrate better performance of the proposed AIC corrections when applied to polynomial regression in comparison to AIC, AIC c and other criteria are presented. Asymptotic justifications for the proposed criteria are provided in the Appendix.}, author = {Seghouane, Abd-Krim}, citeulike-article-id = {5319135}, citeulike-linkout-0 = {http://dx.doi.org/10.1016/j.sigpro.2009.06.010}, citeulike-linkout-1 = {http://linkinghub.elsevier.com/retrieve/pii/S0165168409002837}, day = {16}, doi = {10.1016/j.sigpro.2009.06.010}, issn = {01651684}, journal = {Signal Processing}, keywords = {aic, bootstrap, eic, model, selection}, month = {January}, number = {1}, pages = {217--224}, posted-at = {2009-09-12 00:22:14}, priority = {2}, title = {Asymptotic bootstrap corrections of AIC for linear regression models}, url = {http://dx.doi.org/10.1016/j.sigpro.2009.06.010}, volume = {90}, year = {2010} }

TY - JOUR ID - Seghouane2010 L3 - citeulike-article-id:5319135 N2 - The Akaike information criterion, AIC, and its corrected version, AIC c are two methods for selecting normal linear regression models. Both criteria were designed as estimators of the expected Kullback–Leibler information between the model generating the data and the approximating candidate model. In this paper, two new corrected variants of AIC are derived for the purpose of small sample linear regression model selection. The proposed variants of AIC are based on asymptotic approximation of bootstrap type estimates of Kullback–Leibler information. These new variants are of particular interest when the use of bootstrap is not really justified in terms of the required calculations. As its the case for AIC c , these new variants are asymptotically equivalent to AIC. Simulation results which illustrate better performance of the proposed AIC corrections when applied to polynomial regression in comparison to AIC, AIC c and other criteria are presented. Asymptotic justifications for the proposed criteria are provided in the Appendix. IS - 1 JF - Signal Processing SN - 01651684 EP - 224 TI - Asymptotic bootstrap corrections of AIC for linear regression models VL - 90 SP - 217 KW - aic KW - bootstrap KW - eic KW - model KW - selection AU - Seghouane, Abd-Krim PY - 2010/01/16/ UR - http://dx.doi.org/10.1016/j.sigpro.2009.06.010 ER -