Bootstrap Methods : A Practitioner's Guide

Rating:
Summary: Essential for Resamplers
Review: Anyone with a serious interest in resampling will want to have this book in his or her library. The reference listings, invaluable for researchers, are the most complete currently available for bootstrappers. The material is accessible to applied workers, with mathematical complexity kept at a minimum. Bootstrapping, when approached in the right way, is a tool of great general testing application and minimal complexity. Chernick's book keeps things clear and accessible. This is the right reference book published at the right time.

Rating:
Summary: A complete account of Bootstrap methods
Review: Michael Chernick ... has written this book which may be considered as a first complete account of the Bootstrap methods since Efron's seminal paper in 1979. Chernick has used the Current Index to Statistics CD ROM, and several other search engines ... to collect all journal papers and books that have any relation with the Bootstrap. The result is a list of 1600 references of which more than 600 are actually cited in his book. Apart from this, I found his book interesting to read, especially his sections on the error rate estimation in two-class discrimination problems leading to the so-called .632 estimator (which is one of the big succes stories for the bootstrap), and the applications of bootstrapping in Kriging, analysis of mixture models, censored data analysis, missing data problems, and Bayesian bootstrapping. Finally he devotes a chapter on situations when Bootstrapping might fail, such as in the case of extreme value estimation.