作者: Norman P. Archer
关键词:
摘要: An improved computational technique has been developed for use in the maximum likelihood estimation of Weibull parameters complete, censored or grouped data, 3-parameter model. To demonstrate technique, a limited set Monte Carlo results are given which compare actual covariance matrix parameter estimates with an approximation to this using negative inverse empirical information matrix. The approximation, suggested by several authors, is examined both and standard 2-parameter particular cases censored, samples sizes.