Modern Regression Methods (Wiley Series in Probability and Statistics)

The most comprehensive book available on state-of-the-art regression methodology, complete with exercises and solutions This combination book and disk set presents the full range of regression techniques available today to practitioners, researchers, and students of this popular and ever-changing field. Featuring a strong data analysis orientation and a more comprehensive treatment of regression diagnostics than is found in other texts, Modern Regression Methods contains a wealth of material assembled here for the first time, including recently developed techniques and some new methods introduced by the author, as well as fresh approaches to standard concepts. With thorough analyses of real-world data sets and many exercises with worked solutions, this unique resource reinforces learning while providing you with crucial hands-on experience in the practical application of skills. The book offers: In-depth treatment of standard regression methods, including diagnostics, transformations, ridge regression, and variable selection techniques A detailed examination of nonlinear regression, robust regression, and logistic regression, including both exact and maximum likelihood approaches for logistic regression New graphical techniques and transformation strategies for multiple regression and a survey of nonparametric regression Experimental designs for regression Minitab macros to facilitate understanding and use of many of the new methods that are presented Amstat News asked three review editors to rate their top five favorite books in the September 2003 issue. Modern Regression Methods was among those chosen.