 Andrew Gillman's book is the Data Scientist's field guide. It's the practitioner's book. Lots of the books that I've reviewed are more theoretical, but this one is something that you can actually use. Why is that? Well, it has code examples, some relevant exercises, and it even dives into some very specific fields. You'll see things like how to handle missing values, how to work with time series data. There's also some bits on causal modeling. Chapter one and two set the stage. Chapter three starts to introduce basic methods, mathematics, and probability. Here he starts with weighted averages, which naturally lead into a discussion on expectations. He also talks a little bit about vectors and matrices and gets the reader comfortable with some of that notation. He then goes into the probability distributions that you might see in other textbooks. One chapter I love in this book is chapter five, which delves into simulations. Simulations are crucial, but often overlooked, and so I was really happy to see this in Gillman's book. With simulations and some code examples and simulations, he then cleverly transitions from simulations to bootstrap methods, illustrating the use of bootstrap for simulating sampling distributions. He also has a little section on some warnings about using the bootstrap. The book flows really nice. This is something I feel like is rare in these kinds of books. I would almost have expected something that was a little bit more like a dictionary, but this book flowed from one topic to the next very well. You can tell that Andrew Gillman's been doing this for a while. So getting back to chapter four, simulation techniques such as resampling, bootstrapping, Monte Carlo methods are thoroughly explored, highlighting the importance of an early familiarity with these concepts. He'll even delve into simulating fake data. Here he emphasizes that models should generate data resembling the fitted data. As a Bayesian, Gillman views model fitting through the lens of understanding the underlying data generation process with simulation being a key tool. Part two finally gets into the meat of regression modeling. In the chapter on regression modeling, standard formulas are presented focusing more on practical application than theory. The book, again, is rich in code here, showing lots of examples and plots, offering a comprehensive view of model visualization. Topics like cross-validation, including leaving out cross-validation, K-fold methods, and then also AIC are covered in chapter 10. Chapter 15 builds on earlier foundations exploring the Poisson and negative binomial regression. He also addresses zero-inflated data. Again, this is just such a unique book. You get to work with these real problems that you get in data and you get to see the code he would use. The book also touches on some more advanced methods like thinking about using regression with the T model or the T distribution. The exercises in the book are approachable, requiring no advanced degree. And the book concludes with Appendix B. Normally what you'll find in the Appendix of a math book is some review of math concepts, which is great. It gets you familiar with the notation they want to use for math concepts you might already be familiar with. But in Andrew Gilman's book in Appendix B, he leaves you with a gift. It's 10 quick tips to improve your regression modeling. I'm not gonna show you all of them in this video. I'll just highlight my favorite one. He encourages the modeler to fit many models, emphasizing that the first model is not sacred and should be continuously improved. In summary, regression and other stories by Andrew Gilman is an essential desk reference for practitioners. It provides a comprehensive understanding of regression, benefiting even those working on non-regression models. This book is a must have for every data scientist enhancing efficiency and proficiency in data utilization. As a personal anecdote, those who understand regression better are better modelers in general. Thanks for watching and we'll see you next time.