No label defined (Q4490261)

The book under review is intended to be used as a text for an introductory course on model based parametric inference in senior undergraduate (honors) or in the first year graduate course. It consists of ten chapters, References, and the Index. Chapter 1, named ``Introduction'', provides a historical perspective, which leads to the frequentist approach where the evaluation of the performance of any inference procedure is based on its sampling distribution. Chapter 2, named ``Sufficient Statistics'', introduces the concept of the Fisher information in a sample about the unknown parameter of interest. Next, the problem of defining and deriving a minimum sufficient statistic is discussed, followed by considering the exponential and the Pitman families of distributions. Chapter 3, named ``Minimum Variance Unbiased Estimation'' (MVUE), dwells on such topics as unbiasedness, BLUE (best linear unbiased estimator), Cramér-Rao inequality and its applications, Rao-Blackwell theorem, completeness and Lehmann-Scheffé theorem, and a necessary and sufficient condition for MVUE. Chapter 4, named ``Simultaneous Estimation of Several Parameters'', presents the matrix optimality criterion, Klebanov-Linnik-Rukhin's theorem, generalized Cramér-Rao inequality, and comparison of optimality criteria. Chapter 5, named ``Consistent Estimators'', discusses such topics as consistency, the method of moments, the method of percentiles, and the problem of choosing between consistent estimators. Chapter 6, named ``Consistent Asymptotically Normal Estimators'' (CAN), contains sections presenting the basic result, the method of moments and percentiles, CAN estimators for a multiparameter case, and the discussion of the method of least squares and the Boscovitch method based on \(L_1\)-norm. Chapter 7, named ``Method of Maximum Likelihood'' (ML), considers ML estimation in the Exponential family, the Cramér family, Cramér-Huzurbazar theorem, solution of likelihood equations, and asymptotically most efficient estimators. Chapter 8, named ``Tests of Hypotheses - I'', presents critical regions and test functions, Neyman-Pearson's lemma and the most powerful test, and the issue of uniformly most powerful tests and their existence. Chapter 9, named ``Tests of Hypotheses - II'', discusses the likelihood ratio test (LRT), LRT for multinomial distributions, large sample tests and their consistency, and the Bayesian approach to hypotheses testing. Chapter 10, named ``Interval Estimation'', considers shortest expected length confidence intervals, large sample confidence intervals, unbiased confidence intervals, and Bayesian and fiducial intervals.