Bootstrap Methods: A Practitioner's Guide

By J. Chimka. 2003

685 citations recorded by the source corpus.

What was originally intended as a companion to Phillip Good’s Permutation Tests has arrived in Bootstrap Methods by Michael R. Chernick, for the Wiley Series in Probability and Statistics. It is a practitioner’s guide through extensive applications that also includes a detailed literature review to motivate advances in the subject. The book begins by generally answering, “What is Bootstrapping?” in Chapter 1 with a background, an introduction and mention of applications. The reader is provided here with a basic explanation of Efron’s bootstrap methods and previews their application to topics such as censored and missing data, subset selection, kriging and p-value adjustment that are discussed in more detail in Chapter 8. An interesting historical perspective and a comprehensive literature review are also presented. Chapters 2 (“Estimation”) and 3 (“Confidence Sets and Hypothesis Testing”) are an update of Efron and Tibshirani’s An Introduction to the Bootstrap. The estimation chapter covers the classification problem and location and dispersion parameters. Section 3.1 deals with nonparametric confidence intervals: the percentile method, bootstrap iteration and the bootstrap method. Section 3.2 connects confidence intervals and hypothesis tests. The clinical trial analysis in Section 3.3 is illustrative (and illustrated). The topic of Chapter 4 is “Regression Analysis.” The bootstrap is shown to help determine confidence and prediction intervals for non-Gaussian error, both in the linear and nonlinear cases. There is also a section on nonparametric models. “Forecasting and Time Series Analysis” are approached in Chapter 5 before the provocative question “Which Resampling Method Should You Use” is posed in Chapter 6. Traditional methods, like the jackknife, crossvalidation and the delta method, are described in this chapter, along with variants of the bootstrap, similarities, differences and preferences. In Chapter 7 (“Efficient and Effective Simulation”) Hall’s bootstrap and the Edgeworth expansion basis for suggesting a number of Monte Carlo iterations are summarized. Practical guidelines are provided. Added special topics in Chapter 8 are determining the number of distributions in a mixture model, process capability indices and point processes. The last chapter highlights one purpose (and another strength) of the book: “to exhibit counterexamples to the consistency of bootstrap estimates so that the reader will be aware of the limitations of the methods.” Cautionary examples are described where bootstrapping fails. The book is chiefly an introduction to resampling procedures that affords the reader just enough theoretical development for a clear and practical understanding of the bootstrap. Still, the statistician with a more advanced theoretical background and/or more experience in bootstrap methods should appreciate the book’s extensive bibliography and chapter-ending historical notes.

Subfields: Artificial Intelligence, Computer Networks and Communications, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/991aa3a93fd8d982bf08ffbb6fe67c8e69e8cc87

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Management Science and Operations Research

Citations in the corpus

Learning Paths to Offshore Outsourcing - From Cost Reduction to Knowledge Seeking (2007, methodology): “Of particular interest is the fact that firms plan to outsource Research & Development to a far greater extent than they do now, with an increase from 11 % of firms to 23 % of firms over the coming three years.”
Data Analysis Using Regression and Multilevel/Hierarchical Models: Understanding and summarizing the fitted models (2006, methodology): “Fienberg (1977) and Agresti (2002) are other useful references.”
Mixed Effects Models and Extensions in Ecology with R (2009, methodology): “Estimation of k depends on the software, but can for example be done in a 2-stage iterative approach (Agresti, 2002).”