摘要
本文提出用Bayes方法对小样本或单次运行的仿真输出进行分析。给出了几种条件下的Bayes方法和伽玛分布置信区间的估计方法。并用M/M/I排队系统对几种方法进行了实例研究。实验结果表明,在处理小样本或单次仿真运行时Bayes方法有明显的优越性和应用前景。
This paper advances that Bayesian methods can be used to analyze small sample or single replication simulation output data, and presents some Bayesian methods and estimating method of confidence interval of Gamma distribution. For these analysis methods, we make some experiment researches by using M / M / 1 queueing system. All results show that Bayesian methods have obvious advantages and application future.
出处
《系统仿真学报》
CAS
CSCD
1993年第3期35-42,47,共9页
Journal of System Simulation
关键词
仿真输出
分析
BAYES方法
随机过程
Discrete-event Simulation output data analysis Bayesian methods Confidence interval
