Bayesian Statistics the Fun Way: Understanding Statistics and Probability with Star Wars, LEGO, and Rubber Ducks Illustrated Edition

It explains all the concepts using very simple, funny and engaging examples! You start off with solving the probability of seeing a UFO to then analyzing a burglary at home! The examples are fantastic and the chapters are just short enough that they keep you engaged throughout. The exercises at the end are very helpful too! I've been enjoying my time learning about Bayesian Statistics

Recently I've gotten more interested into how A/B testing works. As a person who hasn't taking statistics in more than 10 years, I decided to find an approachable book that wouldn't leave me flustered. This book did just the trick. This entire book can be read within a few hours. It doesn't skimp on the mathematics but explains a lot of it via intuition. The example problems it uses are sure to keep your interest. They are original and not the least bit boring. At the end of each chapter you get to practice your skills by writing a few, albeit simple, R programs to test your knowledge. There was only one part of this book that made me go "I wish he would've explained this a bit more" which is the equation for the beta distribution. He seems to gloss over it's construction but covers how to use it, which isn't too bad since R will do the leg work for you. After reading this book I walked away with an understanding of how A/B testing software works such as launch darkly. I also have started to crack open my probability and statistics book from when I was in college and it felt much less dauting this time around. If you want to learn about parameter estimation, hypothesis testing the Bayesian way, you can't find a better intro than this one. Nostarch is constantly producing hits and this one is no exception.

Throughout chapters 3 through 5 there’s no mention of ‘independence’. He just says P(A,B) equals P(A)xP(B). I find myself mentally shouting “what if A and B are not independent?” He doesn’t mention that until chapter 6. He could at least say “If A and B are independent events then P(A,B) = P(A) times P(B); we’ll discuss independence in chapter 6.” In his example of taking a bus or a train, the events of the bus being late and the train being late ARE dependent, since if the bus is late lots of people might crowd into the train, and make it late. I might add some more to this review later.

I read this book and it walks through things in a fun and entertaining and very relatable manner. If you are new to Bayes (like I am), I would recommend writing out the definitions and keeping a journal, it makes it very easy to go back and review Beta, and some of the definitions of priors when you get to the C3PO chapter. I found myself rewriting out some of the prior chapters info as new concepts were added. This is not the kind of book you want to read without working some of the math along the way. I would say if you have a command of some of the beginning concepts in high school algebra or took stat (even better), you will be fine understanding the concepts.Artwork was beautiful, concepts were fun to read and kept my interest. Book arrived in 1 day, fast ship!Great author! Great book, highly recommend for those who are interested in the subject matter. Cannot wait for the next one.

The focus of my professional work is the development and analysis of trading systems. My constant challenge is understanding, monitoring, and measuring the relationships between the model, the data it processes, the target it predicts, and the results it produces.Managing trading systems requires comparing real-time results with benchmarks, estimating risks, and adjusting position size. Being able to estimate confidence intervals is critically important. Bayesian analysis is the primary technique I use.Mr. Kurt clearly explains how individual data values form distributions, and how the shape and metrics of the distributions change as new data is observed. He outlines the theory underlying Bayesian analysis, carefully identifying each of the components and explaining how each is determined. He describes calculation of credible intervals of confidence for means, and probabilities that the means of distributions are different. The descriptions of techniques are clear and presented with a minimal amount of math.This book is an excellent introduction to the techniques used in modeling and managing trading systems.

I purchased this book as an easy-to-read introduction to learn Bayesian statistics for scientific data analysis. The text is fluid, and the book doesn't get too wordy given its mathematical content. I would recommend this book for first time students in statistics or people stepping into data science.One of the things I dislike about mathematics/statistics textbooks is the jargon and the ambiguous symbols, which this books does an excellent job of avoiding. Be warned though, this book is about a technical subject, and there are a lot of mathematical equations, which thankfully are walked through in the text.The author uses a lot of intuitive examples to illustrate the mathematics. However, the book is rather brief, and so this book is better at accompanying, rather than replacing, statistics textbooks.

The book was able to provide a clear and down-to-earth explanation of priors and likelihood using a single/Bernoulli parameter. Before reading this book, I had a hard time differentiating priors from likelihood/data. I always thought that the data from observations, quasi-experiments and simulation was all I had in working with the numbers; it never occurred to me that bringing in seasoned or experienced judgment through priors could formally and mathematically improve the probability of whatever I was estimating. Now, I understand why the frequentist and the Bayesian views clash. It was illuminating to have the tools to counter observations and data when you have other evidence to help in calibrating the final probability or posterior. Kudos to Will Kurt.