Bootstrap of LAD Estimate in Infinite Variance AR(1) Processes

  • Kang, Hee-Jeong (Department of Statistics, Chonbuk Naitonal University, Chonju, Chonbuk 561-756)
  • 발행 : 1997.09.01

초록

This paper proves that the standard bootstrap approximation for the least absolute deviation (LAD) estimate of .beta. in AR(1) processes with infinite variance error terms is asymptotically valid in probability when the bootstrap resample size is much smaller than the original sample size. The theoretical validity results are supported by simulation studies.

키워드

참고문헌

  1. SIAM Journal on Scientific and Statistical Computing v.1 Least Absolute Deviations Curve-Fitting Bloomfield, P.;Steiger, W. L.
  2. The Annals of Statistics v.16 no.4 Edgeworth correction by bootstrap in autoregressions Bose, A.
  3. Stochastic Processes and their Applications v.40 M-estimation for autoregressions with infinite variance Davis, R. A.;Knight, K.;Liu, J.
  4. Annals of Probability v.13 Limit theory for moving averages of random variables with regularly tail probabilities Davis, R. A.;Resnick, S. I.
  5. An Introduction to Probability Theory and Its Applications(2nd Ed.) v.Ⅱ Feller, W.
  6. Journal of Applied Probability v.16 Least absolute deviation estimates in autoregression with infinite variance Gross, S.;Steiger, W. L.
  7. Journal of the Korean Statistical Society v.26 no.1 On Asymptotic Properties of Bootstrap for Autoregressive Processes with Regularly Varying Tail Probabilities Kang, H. J.
  8. Regular Variation and Point Processes Extreme Values Resnick, S. I.