Empirical Bayes - Some Tools, Rules, and Duals

Empirical Bayes methods as envisioned by Herbert Robbins are becoming an essential element of the statistical toolkit. In Empirical Bayes: Tools, Rules, and Duals, Roger Koenker and Jiaying Gu offer a unified view of these methods. They stress recent computational developments for nonparametric estimation of mixture models, not only for the traditional Gaussian and Poisson settings, but for a wide range of other applications. Providing numerous illustrations where empirical Bayes methods are attractive, the authors give a detailed discussion of computational methods, enabling readers to apply the methods in new settings.

List of contents

1. Introduction; 2. The empirical bayes paradigm; 3. Parametric empirical bayes methods; 4. Nonparametric empirical bayes methods; 5. Mixture models; 6. Heteroscedastic gaussian mixture models and longitudinal data; 7. Ranking and selection as compound decisions; 8. Empirical bayes methods for survival analysis; 9. Mixture models for categorical data; 10. Random coefficient models; 11. Inference for empirical bayes procedures; 12. Computational considerations; 13. Conclusion; Bibliography; Subject Index; Names Index.

About the author

Roger Koenker is an Honorary Professor of Economics at University College London. He was a member of the technical staff of Bell Labs and Professor of Economics and Statistics at the University of Illinois. He is a fellow of the Econometric Society, the ASA and the IMS.

Product details

• Authors: Jiaying Gu, Koenker Roger
• Publisher: Cambridge Academic
• Languages: English
• Product format: Hardback
• Release: 31.03.2026
• EAN: 9781009275361
• ISBN: 978-1-009-27536-1
• Weight: 500 g
• Illustrations: Worked examples or Exercises
• Series: Econometric Society Monographs
• Subjects: Social sciences, law, business > Business > General, dictionaries; BUSINESS & ECONOMICS / Econometrics, Economic statistics, Microeconomics, Econometrics, Econometrics and economic statistics, For graduate / post-graduate and equivalents