Comparison of Parameter Estimation for Weibull Distribution

  • Wang, Fu-Kwun (Department of Industrial Engineering and Management National Taipei University of Technology) ;
  • J. Bert Keats (Department of Industrial Engineering Arizona State University, U.S.A.) ;
  • B. Y. Leu (Department of Industrial Engineering and Management Ming-Hsin University of Science & Technology, Taiwan)
  • Published : 2003.03.01

Abstract

This paper represents the first comprehensive comparison of the Newton-Raphson's method and Simple Iterative Procedure (SIP) in the maximum likelihood estimation of the two-parameter Weibull distribution. Computer simulation is employed to compare these two methods for multiply censored, singly censored data (Type I or Type Ⅱ censoring) and complete data. Results indicate the Newton-Raphson's with the Menon's estimated value, as an initial point remains the effective iterative procedure for estimating the parameters.

Keywords