Title（题目）：随机估计和VDR检验
Time（时间）：2013-04-18(星期四) 16:00-17:00
Place（地点）：中国人民大学统计学院（明德主楼1031教室）
Abstract（摘要）：In this paper the concept on randomized estimation of one dimensional parameter is generalized to that of vector parameter. The randomized estimator of vector parameter is defined by using both pivot of the parameters and pivot random vector which has the same distribution function as that of pivot. It root in Fisher's fiducial probability density, but it is still a frequentist concept. The VDR test is constructed by using test variate which is defined though both randomized estimation and its p.d.f. We apply VDR test to the hypothesis on equal means without condition of equal variances and on homogeneity of variances for several normal distributions. The VDR test power is simulated and compared with known tests. The results show that VDR tests are good. The VDR test is a universal method, could be applied to many problems and the classical results can be derived by using VDR test.