摘要
Phase-Type(PH)分布具有良好的通用性和易于进行解析处理的特性,因此PH分布已经成为随机模型解析分析的重要工具。PH分布的数据拟合方法是应用PH分布的基础,分析了PH分布数据拟合的矩估计方法、极大似然估计方法及PH分布拟合重尾分布数据方法的研究现状,并讨论了PH分布数据拟合方法进一步的研究方向。
Phase-Type distributions have gained widespread attention two primary reasons being their denseness and algorithmic tractability. in the area of stochastic modeling, Approximating observation data by phase-type distributions is the first step of using Phase-Type distributions in analytical modeling. First, the status of PH distribution data fitting method is analyzed from three aspects： moment match, maximum likelihood estimation and heavy-tailed distribution data fitting method. Finally, the future development direction of PH distribution data fitting method is discussed.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2008年第11期2167-2174,共8页
Systems Engineering and Electronics
关键词
PH分布
数据拟合
矩估计
极大似然估计
重尾分布
Phase-Type distribution
data fitting
moment match
maximum likelihood estimation
heavytailed distribution
