Geometric分布与寿命分布复合族参数估计的MM算法

MM Algorithm for Parameter Estimation of the CompoundClass of Geometric and Lifetime Distributions

MM算法是处理优化问题的有效工具,应用广泛的EM算法也是MM算法的一个特例.将MM算法引入到Geometric分布与寿命分布复合族(如Exponential-Geometric分布,Weibull-Geometric分布,Pareto-Geometric分布和Rayleigh-Geometric分布)中以求解参数的极大似然估计.随机模拟结果表明MM算法在Geometric分布与寿命分布复合族参数估计中具有优良的性能.

The MM algorithm is an effective tool to deal with optimization problems,and the widely used EM algorithm is also a special case of the MM algorithm.In this paper,the MM algorithm is introduced into the compound class of geometric and lifetime distributions(such as Exponential-Geometric distribution,Weibull-Geometric distribution,Pareto-Geometric distribution and Rayleigh-Geometric distribution)to solve the maximum likelihood estimation of the parameters.The random simulation results show that the MM algorithm has excellent performance in the parameter estimation of the compound class of geometric and lifetimes distributions.