A Practitioner’s Guide to Resampling for Data Analysis, Data Mining, and Modeling
Distribution-free resampling methods—permutation tests, decision trees, and the bootstrap—are used today in virtually every research area. A Practitioner’s Guide to Resampling for Data Analysis, Data Mining, and Modeling explains how to use the bootstrap to estimate the precision of sample-based estimates and to determine sample size, data permutations to test hypotheses, and the readily-interpreted decision tree to replace arcane regression methods.
- Each chapter contains dozens of thought provoking questions, along with applicable R and Stata code
- Methods are illustrated with examples from agriculture, audits, bird migration, clinical trials, epidemiology, image processing, immunology, medicine, microarrays and gene selection
- Lists of commercially available software for the bootstrap, decision trees, and permutation tests are incorporated in the text
- Access to APL, MATLAB, and SC code for many of the routines is provided on the author’s website
- The text covers estimation, two-sample and k-sample univariate, and multivariate comparisons of means and variances, sample size determination, categorical data, multiple hypotheses, and model building
Statistics practitioners will find the methods described in the text easy to learn and to apply in a broad range of subject areas from A for Accounting, Agriculture, Anthropology, Aquatic science, Archaeology, Astronomy, and Atmospheric science to V for Virology and Vocational Guidance, and Z for Zoology.
Practitioners and research workers and in the biomedical, engineering and social sciences, as well as advanced students in biology, business, dentistry, medicine, psychology, public health, sociology, and statistics will find an easily-grasped guide to estimation, testing hypotheses and model building.
Table of Contents
Wide Range of Applications. Estimation and the Bootstrap. Software for Use with the Bootstrap and Permutation Tests. Comparing Two Populations. Multiple Variables. Experimental Design and Analysis. Categorical Data. Multiple Hypotheses. Model Building. Classification. Restricted Permutations. References. Appendix A: Basic Concepts in Statistics. Appendix B: Proof of Theorems. Author Index. Subject Index.
Phillip Good is the author of 18 novels, 625 popular articles in magazines and newspapers, scholarly articles in the fields of astrophysics, biology, biostatistics, computer science, probability, and statistics, and nine statistical texts including Applying Statistics in the Courtroom: A New Approach for Attorneys and Expert Witnesses, Chapman Hall, London, 2001. ISBN 1-58488-271-9, and Managers' Guide to the Design and Conduct of Clinical Trials, Wiley, NY, 2002 (2nd edition, 2006).