A general framework for frequentist model averaging
- 5 Downloads
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.
Keywordsasymptotic distribution bias variance trade-off local mis-specification model averaging estimators optimal weight selection
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.
- 10.Hastie T, Tibshirani R, Friedman J. The Elements of Statistical Learning: Data Mining, Inference and Prediction. New York: Springer-Verlag, 2009Google Scholar
- 24.Mitra P. Topics in model averaging & toxicity models in combination therapy. PhD Thesis. New Brunswick: Rutgers University, 2015Google Scholar
- 30.Van der Vaart A W. Asymptotic Statistics, Volume 3. Cambridge: Cambridge University Press, 2000Google Scholar