Abstract
Model selection strategies have been routinely employed to determine a model for data analysis in statistics, and further study and inference then often proceed as though the selected model were the true model that were known a priori. Model averaging approaches, on the other hand, try to combine estimators for a set of candidate models. Specifically, instead of deciding which model is the `right' one, a model averaging approach suggests to fit a set of candidate models and average over the estimators using data adaptive weights. In this paper we establish a general frequentist model averaging framework that does not set any restrictions on the set of candidate models. It broadens the scope of the existing methodologies under the frequentist model averaging development. Assuming the data is from an unknown model, we derive the model averaging estimator and study its limiting distributions and related predictions while taking possible modeling biases into account. We propose a set of optimal weights to combine the individual estimators so that the expected mean squared error of the average estimator is minimized. Simulation studies are conducted to compare the performance of the estimator with that of the existing methods. The results show the benefits of the proposed approach over traditional model selection approaches as well as existing model averaging methods.
Article PDF
Similar content being viewed by others
Avoid common mistakes on your manuscript.
References
Billingsley P. Probability and Measure. Chichester: John Wiley & Sons, 2008
Buckland S T, Burnham K P, Augustin N H. Model selection: An integral part of inference. Biometrics, 1997, 53: 603–618
Claeskens G, Hjort N L. Model selection and model averaging. J Math Psych, 2008, 44: 92–107
Danilov D, Magnus J R. Forecast accuracy after pretesting with an application to the stock market. J Forecast, 2004, 23: 251–274
Danilov D, Magnus J R. On the harm that ignoring pretesting can cause. J Econometrics, 2004, 122: 27–46
Draper D. Assessment and propagation of model uncertainty. J R Stat Soc Ser B Stat Methodol, 1995, 57: 45–97
Gao Y, Zhang X, Wang S, et al. Model averaging based on leave-subject-out cross-validation. J Econometrics, 2016, 192: 139–151
Giles D E A, Lieberman O, Giles J A. The optimal size of a preliminary test of linear restrictions in a misspecified regression model. J Amer Statist Assoc, 1992, 87: 1153–1157
Hansen B E. Least squares model averaging. Econometrica, 2007, 75: 1175–1189
Hastie T, Tibshirani R, Friedman J. The Elements of Statistical Learning: Data Mining, Inference and Prediction. New York: Springer-Verlag, 2009
Hjort N L, Claeskens G. Frequentist model average estimators. J Amer Statist Assoc, 2003, 98: 879–899
Hjort N L, Claeskens G. Focused information criteria and model averaging for the Cox hazard regression model. J Amer Statist Assoc, 2006, 101: 1449–1464
Hoeting J A, Madigan D, Raftery A E, et al. Bayesian model averaging. Statist Sci, 1999, 14: 121–149
Holland P W, Welsch R E. Robust regression using iteratively reweighted least-squares. Comm Statist Theory Methods, 2007, 6: 813–827
Hurvich C M, Tsai C-L. Regression and time series model selection in small samples. Biometrika, 1989, 76: 297–307
Hurvich C M, Tsai C-L. Bias of the corrected AIC criterion for underfitted regression and time series models. Biometri- ka, 1991, 78: 499–509
Karagrigoriou A, Lee S, Mattheou K. A model selection criterion based on the BHHJ measure of divergence. J Statist Plann Inference, 2009, 139: 228–235
Lehmann E L. Elements of Large-Sample Theory. Springer Texts in Statistics. New York: Springer-Verlag, 1999
Lehmann E L, Casella G. Theory of Point Estimation, 2nd ed. Springer Texts in Statistics. New York: Springer-Verlag, 1998
Liang H, Zou G, Wan A T K, et al. Optimal weight choice for frequentist model average estimators. J Amer Statist Assoc, 2011, 106: 1053–1066
Lien D, Shrestha K. Estimating the optimal hedge ratio with focus information criterion. J Futures Markets, 2005, 25: 1011–1024
Madigan D, Raftery A E, York J C, et al. Strategies for graphical model selection. In: Selecting Models from Data: Artificial Intelligence and Statistics IV. Lecture Notes in Statistics, vol. 89. New York: Springer, 1994, 91–100
Magnus J R, Wan A T K, Zhang X. Weighted average least squares estimation with nonspherical disturbances and an application to the Hong Kong housing market. Comput Statist Data Anal, 2011, 55: 1331–1341
Mitra P. Topics in model averaging & toxicity models in combination therapy. PhD Thesis. New Brunswick: Rutgers University, 2015
Pesaran M H, Schleicher C, Zaffaroni P. Model averaging in risk management with an application to futures markets. J Empir Finance, 2009, 16: 280–305
Posada D, Buckley T R. Model selection and model averaging in phylogenetics: Advantages of Akaike information criterion and Bayesian approaches over likelihood ratio tests. Systematic Biology, 2004, 53: 793–808
Raftery A E, Madigan D, Hoeting J A. Bayesian model averaging for linear regression models. J Amer Statist Assoc, 1997, 92: 179–191
Stamey T A, Kabalin J N, Ferrari M, et al. Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate, IV: Anti-androgen treated patients. J Urol, 1989, 141: 1088–1090
Thursby J G, Schmidt P. Some properties of tests for specification error in a linear regression model. J Amer Statist Assoc, 1977, 72: 635–641
Van der Vaart A W. Asymptotic Statistics, Volume 3. Cambridge: Cambridge University Press, 2000
Wan A T K, Zhang X, Zou G. Least squares model averaging by Mallows criterion. J Econometrics, 2010, 156: 277–283
Wei Y, McNicholas P D. Mixture model averaging for clustering and classification. Adv Data Anal Classif, 2015, 22: 197–217
Zhang X, Wan A T K, Zhou S Z. Focused information criteria, model selection and model averaging in a tobit model with a non-zero threshold. J Bus Econom Statist, 2012, 30: 132–142
Zhang X, Yu D, Zou G, et al. Optimal model averaging estimation for generalized linear models and generalized linear mixed-effects models. J Amer Statist Assoc, 2016, 111: 1775–1790
Zhang X, Zou G, Carroll R J. Model averaging based on Kullback-Leibler distance. Statist Sinica, 2015, 25: 1583–1598
Zhang X, Zou G, Liang H. Model averaging and weight choice in linear mixed-effects models. Biometrika, 2014, 101: 205–218
Acknowledgements
The work was supported by National Science Foundation of USA (Grant Nos. DMS- 1812048, DMS-1737857, DMS-1513483 and DMS-1418042) and National Natural Science Foundation of China (Grant No. 11529101). This article is a work developed based on the thesis of the first author. The authors wish to use this article to celebrate Professor Lincheng Zhao's 75th birthday and his tremendous and long lasting contribution to statistical research and education in China and around the world. The authors also thank two referees for their valuable suggestions and comments that have helped improve the paper substantially.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Mitra, P., Lian, H., Mitra, R. et al. A general framework for frequentist model averaging. Sci. China Math. 62, 205–226 (2019). https://doi.org/10.1007/s11425-018-9403-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11425-018-9403-x
Keywords
- asymptotic distribution
- bias variance trade-off
- local mis-specification
- model averaging estima- tors
- optimal weight selection