数理统计期末考题答案.docVIP

Let be a random sample from the distribution , , , . ( 8 %) Find the method of moment estimates of and . ( 7 %) Find the MLE of , assuming is known. ( 7 %) Giving , find the Cramer-Rao lower bound of estimates of . ( 8 %) Giving , find the UMVUE of . Suppose that are iid ~, . Let . ( 5 %) Show that is a sufficient statistic for . ( 5 %) Let . Show that is an unbiased estimate of . (10%) Find the UMVUE of . Let be a random sample from a ,, distribution. Consider testing vs . (10%) Find a UMP level test, . ( 7 %) For , the test rejects , if . Find the power function of the test. ( 8 %) For , the test rejects , if . Evaluate the size and the power of the test. (10%) Let be iid distribution, and let the prior distribution of be a distribution, , . Find the posterior distribution of . Let be a random sample from an exponential distribution with mean ,. ( 5 %) Show that is a sufficient statistic n for . ( 5 %) Show that the Poisson family has a monotone likelihood ratio, MLR. ( 5 %) Find a UMP level test of vs by the Karlin-Rubin Theorem shown below. [Definition] A family of pdfs or pmfs has a monotone likelihood ratio, MLR, if for every , is a monotone function of . [Karlin-Rubin Theorem] Suppose that is a sufficient statistic for and the pdfs or pmfs has a non-decreasing monotone likelihood ratio. Consider testing vs . A UMP level test rejects if and only if , where . 數理統計期末考試試題答案 (a) Since and , Let and , . Furthermore, , , The MME of .and are , (b) Let . Furthermore, , So, is the MLE of . (c) CRLB = (d) Since , is an unbiased estimate of , and CRLB, is the UMVUE of . [Or] Given , is an exponential family in . is a sufficient statistic for . Since is an unbiased estimate of and a function of sufficient statistics , by Rao-Blackwell Theorem, is the UMVUE of . (a) Let and . By factorization theorem, is a sufficient statistic for . [Or] is an exponential family is a sufficient statistic. (