A Student’s Guide to Bayesian Statistics

As an engineer we are taught almost exclusively from the frequentist paradigm, and I felt that I needed to self-teach Bayesian statistics if I wanted to get into the realm of forecasting and general modelling. Hours in the student library trawling through texts only came up with extremely dense material. I ended up turning to youtube for some introductory lessons and stumbled across the authors fantastic channel (It can be found by typing “Ben Lambert” into the search bar). The videos were clear, concise and very informative. I found that they were meant to be consumed alongside this text, which I promptly purchased. The quality of this text cannot be stressed enough. Humorous and engaging, it reads like a novel and explains like top quality lecture notes. It walks you through the mathematical fundamentals of Bayesian stats, terminating with a comprehensive guide to Conjugate and Uninformative priors. The final third of the text is dedicated to computational (read: practically used) Bayesian stats, covering topics including Stan and Hierarchical Modelling. The author recommends using R for the problem sets, but I managed with PyStan interface fine so that shouldn’t be a concern. I will certainly have this text on my desk for the foreseeable future as I get more comfortable with solving these problems regularly. The next step is “Bayesian Data Analysis” by the legend Gelman himself, which from what I have seen is prohibitively dense without a first studying a text such as this one.

As a bioinformatician specializing in applied statistics, I found this book to be an excellent introduction to Bayesian Statistics for students and researchers with a non-mathematical background. The concepts and ideas are introduced with exceptional clarity. For those aiming to learn more about Bayesian statistics on their own, this and McElreath’s book will be essential stepping stones towards generating an understanding of more advanced textbooks and modern literature in applied Bayesian statistics. Top aspects: + Clear pedagogical text layout; concise introductions and chapter summaries + Good coverage of topics relevant to applied Bayesian analysis work (probability theory interpretations, distributions, model evaluation, model fitting algorithms, hierarchical & regression modeling ) + Excellent introduction to Bayesian model fitting algorithms (Metropolis Hastings, Gibbs and HMC) from a conceptual angle + Excellent example figures + Brief introductory chapter on stan for those who haven’t used it before It is worth noting that the book has rather few applied statistics examples; the focus of the book clearly lies more on explaining the theory. On this end McElreath’s, Krushke’s and Hilbe’s books are more extensive and contain more examples with stan code. For those wanting to get started with Bayesian statistics I thus recommend getting one of these references and their corresponding R & stan code as a supplement to Lambert's book. I own both a hardcopy and digital copy of the book; the digital version (google play) is overall quite good but for some reason contains rather blurry pictures and weirdly formatted equations. The hardcopy version has excellent resolution figures and nicely formatted equations. I highly recommend getting a hardcopy of this book.

説明が丁寧で入門書としては最適だいと思います

I needed a refresher on Bayesian stats. While browsing/watching related youtube videos I have stumbled upon the author's channel. This is how I learned about 'A Student's Guide to Bayesian Statistics'. In retrospect I cannot believe this was such a random sequence of events, since the "Student's Guide' has now become my personal favourite text book I ever had. The material is presented in a very clear way, it builds up from simple examples to more complicated ones. To go one step further the book offers a few pretty advanced problems to work out (there are answers/solutions available on-line too on the author's web-site). In theory all the text books should be like that, but in practice it is not all that frequent, especially when it comes to any sort of applied math. It was important for me that the text has no insane logic gaps along the lines of 'now, obviously' on which I tend to hang up particularly badly. The illustrations are fantastic too. I feel if the author ever decides to write another book on any subject even remotely relevant to my fields of interest - I will buy it without hesitation.

El libro es muy claro, en un tema poco intuitivo, con buenos ejemplos de la vida real vista desde la perspectiva bayesiana. Para quien quiera adentrarse en la estadística de Bayes desde el principio, esta muy bien. Pero quería comentar que por fin, un libro que verdaderamente da lo que promete: recursos on line del libro, muy completos. Hay videos explicativos de todos los capítulos, y conceptos. Los problemas están todos resueltos. No se suele encontrarse tanto apoyo al lector. Lo recomiendo mucho.