Introduction to Empirical Bayes: Examples from Baseball Statistics

by David Robinson

Learn to use empirical Bayesian methods for estimating binomial proportions, through a series of intuitive examples drawn from baseball statistics. These methods are effective in estimating click-through rates on ads, success rates of experiments, and other situations common in modern data science.

You'll learn both the theory and the practice behind empirical Bayes, including computing credible intervals, performing Bayesian A/B testing, and fitting mixture models. Each example is accompanied with visualizations to demonstrate the mathematical concepts, as well as R code that can be adapted to analyze your own data.

Genres: Mathematics

242 pages, Kindle Edition

Published February 8, 2017

Reviews

[Mark · Rating 4 out of 5]

A very clear and practical introduction to empirical Bayes estimation including R code and all based on simple baseball data. The author first explains several topics of gradually increasing complexity using base R and the tidyverse and then summarizes what he has covered and shows how it can be done with the ebbr package.

[Benjamin Schneider · Rating 5 out of 5]

This is really an excellent book for getting an introduction to Empirical Bayes (and Bayes in general!). The motivation and intuition behind the applications are all quite clear, and the graphics throughout the book are excellent.

[Daniel Hadley · Rating 5 out of 5]

I hope to write a fuller review in the future, but I want to say that this was an enjoyable and insightful introduction to EB. I really liked the framework of starting with simple models and working up through more complex analysis. This allowed me to follow the thread as though I were looking over the author's shoulder while he analyzed a dataset. I gained as much from the process as I did from the clear, *tidy* code.