摘要
在利用Bayes小样本理论对导弹飞行试验可靠性指标进行评定时,一个核心问题是如何从先验信息中得到参数的先验分布。结合具体的算例,给出了两种在实践中较为好用的先验分布超参数的确定方法,分别是分位数法和专家打分法。
When appraising the reliability indexes in the missile flying test with Bayes small sample theory, an essential problem is how to work out the prior distribution of parameters from prior information. Combining with concrete examples, two good methods for establishing prior super parameters have been introduced. The two methods are quantile method and expert scoring method.
出处
《电子产品可靠性与环境试验》
2013年第5期14-16,共3页
Electronic Product Reliability and Environmental Testing
关键词
贝叶斯
小样本
先验分布
超参数
Bayes
small sample
prior distribution
super parameters
