摘要
引入 Bootstrap重采样技术 ,根据少量历史数据获得合格率分布的非参数化数值解 ,作为先验条件与当前样本分布相结合 ,通过贝叶斯方法对尺寸质量参数进行估计。避免传统的参数化先验分布所带来的人为假设 ,提出的方法更具有客观性 。
A Bootstrap resampling method was implemented to build the nonparametric prior empirical distribution of pass-rate, then Bayesian statistical algorithm is used to infer the pass-rate by combining prior empirical distribution with the currently observed distribution. The discrete form of the Bayesian inference for pass-rate is derived, and a Matlab program is developed to perform Bootstrap resampling and Bayesian calculation. A statistical simulation example is also given to verify the effectiveness and precision of this statistical inference.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2003年第4期347-349,共3页
China Mechanical Engineering
基金
国家自然科学基金资助项目 (50 1 750 71 )
