The Statistical Analysis of Parallel System for Geometric Distribution Under the Full Sample

摘要

Considering the parallel system of two units with geometric distribution, the moment estimator and maximum likelihood estimator of parameters are given under the full sample. The precisions of the point estimators are investigated by a large number of Monte-Carlo simulations and it can be concluded that the maximum likelihood estimator is better than the moment estimator. Furthermore, the approximate interval estimator of parameters is proposed and the precision of the interval estimators is also investigated by Monte-Carlo simulations.