Bootstrap confidence bands for shrinkage estimators

作者: Camilla Kazimi , David Brownstone

DOI: 10.1016/S0304-4076(98)00037-2

关键词: Confidence and prediction bandsBootstrapping (electronics)Confidence intervalEdgeworth seriesEconometricsProbability distributionStatisticsMathematicsMonte Carlo methodPercentileEstimator

摘要: Empirical application of shrinkage estimators has been limited by the inability to estimate confidence intervals. We investigate band based on Edgeworth expansions and bootstrapping. Our Monte Carlo results suggest that both Efron's bias correction method with acceleration simple percentile bootstrap methods generate reasonable bands. Approximations performed poorly. then use a single bands for predictions GNP growth rates from leading-indicators model. study shows perform well enough support empirical applications estimators.

参考文章(24)
H.D. Vinod, 23 Bootstrap methods: Applications in econometrics Handbook of Statistics. ,vol. 11, pp. 629- 661 ,(1993) , 10.1016/S0169-7161(05)80058-6
W. James, Charles Stein, Estimation with Quadratic Loss Proceedings of the Fourth Berkeley Symposium on Mathematical Statistics and Probability, Volume 1: Contributions to the Theory of Statistics. pp. 443- 460 ,(1992) , 10.1007/978-1-4612-0919-5_30
Rudolf Beran, Prepivoting Test Statistics: A Bootstrap View of Asymptotic Refinements Journal of the American Statistical Association. ,vol. 83, pp. 687- 697 ,(1988) , 10.1080/01621459.1988.10478649
Robert J Tibshirani, Bradley Efron, An introduction to the bootstrap ,(1993)
H.D. Vinod, Double bootstrap for shrinkage estimators Journal of Econometrics. ,vol. 68, pp. 287- 302 ,(1995) , 10.1016/0304-4076(94)01639-H
Arnold Zellner, Chansik Hong, Forecasting international growth rates using Bayesian shrinkage and other procedures Journal of Econometrics. ,vol. 40, pp. 183- 202 ,(1989) , 10.1016/0304-4076(89)90036-5
David Brownstone, Bootstrapping improved estimators for linear regression models Journal of Econometrics. ,vol. 44, pp. 171- 187 ,(1990) , 10.1016/0304-4076(90)90078-8