New confidence interval estimation method for parameters of J-M model

Bayes statistics is applied to confidence interval estimation for parameters of J-M model. Firstly, Prior distribution of J-M model's parameters based on Jeffrey principle is obtained to get the posterior distribution density which is used to achieve the point estimation of parameters based on Bayes theorem. Bayes interval estimation method is developed consequently. Finally, three interval estimation methods respectively based on maximum likelihood, least square and Bayes theorem are applied to several open failure data sets. The results show that the precision, stability and applicability of interval estimation method based on Bayes theorem is the best in these three methods.