EMPIRICAL LIKELIHOOD INFERENCE FOR LOGISTIC EQUATION WITH RANDOM PERTURBATION

在线阅读 下载PDF 导出详情
摘要 Empiricallikelihood(EL)combinedwithestimatingequations(EE)providesamodernsemi-parametricalternativetoclassicalestimationtechniquessuchasmaximumlikelihoodestimation(MLE).ThispapernotonlyusesclosedformofconditionalexpectationandconditionalvarianceofLogisticequationwithrandomperturbationtoperformmaximumempiricallikelihoodestimation(MELE)forthemodelparameters,butalsoproposesanempiricallikelihoodratiostatistic(ELRS)forhypothesesconcerningtheinterestingparameter.MonteCarlosimulationresultsshowthatMELEandELRSprovidecompetitiveperformancetoparametricalternatives.
机构地区 不详
出版日期 2014年02月12日(中国期刊网平台首次上网日期,不代表论文的发表时间)