Choice of Model for Reliability Studies and Related Topics

摘要

In chapter 1 the author reviews some of the reasons the Weibull distribution has been chosen as a model for studies in several fields. In chapter 2 the likelihood ratio test for discriminating between two distributions with unknown location and scale parameters is considered. The case of selecting between the Weibull and the Log-Normal may be regarded as selecting between two location and scale parameter distributions and this particular problem is presented in chapter 3. Chapter 4 presents the sample sizes needed in order to select the better of two Weibull populations. Chapter 5 presents tables needed in order to obtain an upper confidence limit on the maximum of m future observations from a Type 1 distribution for maximum. (Author)