Bayesian methods : an analysis for statisticians and interdisciplinary researchers /

Leonard (Univ. of Edinburgh) and Hsu (Univ. of California, Santa Barbara) provide excellent, thorough, and very insightful coverage of Bayesian methods used primarily by interdisciplinary researchers and statisticians. The authors cite two philosophies of statistical thinking, the Bayesian (prior and posterior approach) and the Fisherian (frequency approach), but they believe that the Bayesian philosophy possesses numerous advantages because it behaves extremely well when simulated by a computer, thus allowing the researcher to make more sound conclusions and decisions. There is in-depth coverage of likelihood and probability, which is woven throughout the book, thus linking both philosophies. The reader should have an extensive background in mathematics and statistics. This book has broad applications for research in the areas of statistics, economics, medicine, genetics, engineering, etc. Some topical coverage includes the Bayesian paradigm, estimating a discrete valued parameter, symmetric loss functions, the Kalman filter, Bayesian forecasting, and posterior mode vectors and Laplacian approximations. Highly recommended for upper-division undergraduate and graduate students as well as faculty and professionals. D. J. Gougeon; University of Scranton