摘要
基于信息论中最大熵的概念,并结合随机加权法的特点,研究了Bayes统计分析方法中先验分布的获取问题,尽量避免受到主观因素影响的先验分布。同时给出了该方法的实现步骤及其在解决小样本条件下可靠性评估中的运用,并进行了仿真验证分析。仿真结果说明了该方法的可行性以及在小子样条件下进行可靠性评估的优越性。
Based on the maximum entropy in the information theory, and the Bayes Bootstrap Methods the problem of how to obtain the prior distribution of the Bayes statistic method without the subjective factor is examined. The steps to carry out the method and its application in the reliability evaluation under small-sampling condition are presented and a simulation is performed for the verification analysis. The simulation result demonstrated the feasibility of the method and the advantages of its application to reliability evaluation under small sample condition.
出处
《电子产品可靠性与环境试验》
2008年第6期62-65,共4页
Electronic Product Reliability and Environmental Testing
关键词
贝叶斯方法
随机加权
最大熵方法
可靠性评估
Bayes method
Bayes Bootstrap
maximum entropy method
reliability evaluation
