With Applications in R (Springer Texts in Statistics)

Well-written book. Clear explanations of statistical concepts and methodologies. along with explanation of how to implement in R. Useful for someone with a medium understanding of statistics and aptitude for learning programming languages, to rapidly increase their knowledge and gain practice.

How often do the authors of the canonical text in a field recognize the need for a comprehensive introduction to the subject? Not often, I think, but here you have it. I just completed the Coursera Data Analysis course (also excellent) where I was encouraged to read and refer to the canonical text but I simply was not prepared for it; I bought it but I could not make use of it.

I have now plowed through every last chapter in this book, 10 chapters in all, doing the R Labs along the way and I feel fully prepared for the online course the authors are teaching at Stanford this January (2014) and also fully prepared to tackle The Elements of Statistical Learning: Data Mining, Inference, and Prediction, which is what I mean by the canonical text.

This book is a gem and I highly recommend it.

Best introduction to machine learning a girl could hope for. I used this many times in my first year of being a data scientist. Everything is broken down for you. There are many good introductory techniques in here, including regularization, random forests, support vector machines, and PCA. Neural networks are noticeably absent. I recommend the paper copy, but also the Stanford online lectures that go with it.

Best introduction to machine learning a girl could hope for. I used this many times in my first year of being a data scientist. Everything is broken down for you. There are many good introductory techniques in here, including regularization, random forests, support vector machines, and PCA. Neural networks are noticeably absent. I recommend the paper copy, but also the Stanford online lectures that go with it.

Professor Hastie and his team have my utmost respect. They have brought data science into the mainstream more than anyone else I know. And this is their first "introductory" book, and the price was certainly right. Was this your "training sample," Dr. Hastie?

The treatment was good, but it would have been better by making sure the reader already knows probability theory fairly well. Or else, don't explain things in those terms. Also, where is the chapter on simulation and issues associated with that (like outliers)?

Overall, a fine piece of work, something that I'm sure will shine in the second edition (not printing, but edition). In the meantime, it sure helped me give a comprehensive tutorial to about three dozen data scientists in India, most of whom were not fully versed in Western culture or language.

The best university text I've ever read. Very, very deep material, presented in a very intuitive way. Use this book with open source R and RStudio, do the exercises, and you will become proficient in R.

Overall: I did not like both the content and quality of the book.

The book is very heavy for the size, because the paper is thick and glossy. As a result it is also very hard to take notes and highlight sentences in it. It also did not feel like it’s new - maybe because of the paper again. It opened very easy and I did not like the smell.

Now, for content - I liked the range of topics covered. It’s a good intro book in that sense. But there are almost no specific details, there is no math to back it up. If there is an occasional equation, it is often not used later or not explained well. Overall, I found the information to be too general to be useful beyond knowing about the method.

This is a wonderful book. It is clear and straight to the point. It involves enough math to provide understanding of what the algorithms are doing, but focuses on application. Every chapter ends with how to use R to do what was just covered! No googling to find a R library needed. I find myself constantly going back to this book for reference. I would recommend this book to anyone interested in machine learning. By the end, you will be able to build models with R. It gives enough of an overview for the reader to discover interesting areas, and purse those areas with other books.

Very poor condition. And there are a few pages binded in wrong order. Very disappointed. I wanted to return the book but it takes time to get another copy. Furthermore, I have returned one copy already as I was not satisfied with that one either.

I have some 14 statistics textbooks on my bookshelf. It is not hard for me to open a book and find pages and pages of equations with variable letters, Greek letters, subscripts, superscripts, and matrix notation, with not a single number and only the lightest explanation in words. This textbook is different. It is distinctly applied and very well explained using plain English. There is sufficient theoretical material here and the authors occasionally point the reader to the more rigorous Elements of Statistical Learning for deeper theoretical understanding of the concepts. The R programming labs at the end of each chapter successfully reinforce the concepts in the chapter readings. With the trend towards online learning, this book has found the right balance of sufficiently technical, skills-based learning and great explanations to fill the gap when lectures are not available.

I agree with many of the reviews here that this is an excellent book. It is my favorite stats book on the bookshelf.

This book it not meant for beginners. You have to have good statistical backgrounds to understand this book.
The writer mostly uses the ISLR package that, we wrote for R. This package is not up to date with the current version of R studio.

This book is available for free online, but I found it difficult to read the PDF for long time on computer/tablet, so purchased the book.

The book does a great job of introducing machine learning from a statistical perspective. You also get to learn R. It covers a lot of ground and covers them quite well. Springer has done a wonderful job as usual with their acid-free paper and excellent editors.

CONS: Chapter 3 is somewhat long and the authors keep referring to figures in previous pages. (I almost gave up on the book when reading chapter 3.) It may be handy to keep a decent intro statistics book like Statistics, 4th Edition handy when reading this book. It would have also been nice had the authors referred where some of the basic formulas could be studied in more detail. Another thing is some of the advanced topics could have been moved to separate chapters or appendices (esp the topics the authors skip in their online course). Chapter 7 is quite advanced, uses esoteric techniques and could have been presented after decision trees (ch 8) and SVMs (ch 9).

If you are a Python dev, then go with Python Machine Learning instead. It is more practical, covers most of the practical topics in ISLR, teaches neural nets and some NLP and is presented from a machine learning practitioner's perspective, rather than a statistician's perspective.

If you are a practitioner and use R, you may also want to get Applied Predictive Modeling as a follow-up book. It is written by the author of the caret R machine learning package and discusses more practical issues.

Outstanding. Of all the books that I have used in my predictive analytics courses, this is by far the best. The text and examples are comprehensible and interesting. The smooth integration with R is very helpful. In my view, the best part of the book are the accompanying videos. Hastie and Tibshirani were engaging and entertaining in the video, which helped me to absorb the material. I've tried to think of a good analogy for Hastie and Tibshirani's videos...and the winner is NPR's Car Talk. It was great watching these gurus in the field chatting through the material as if they were out for beers with a friend interested in learning the subject matter. Their humility and humor were engaging. Finally, it is great that they make the pdf of the book, videos, and slides all available for free. Bravo.
Search the web for the free videos.

This book is probably not for experts (Springer has a more advanced book on Statistical Learning) but it's great for people who intermediate intermediate R users with a reasonable grasp of regression. The examples are easy to follow and the explanations are clear. This book is meant as a guide to IMPLEMENTING machine learning techniques in R. It does not cover the theory or the math behind the methods, nor does it offer proofs.

If you need a practical guide to implementing statistical learning in R, this is an excellent choice. Very understandable and accessible.

This book introduces the important topics regarding Classification and Regression in unified fashion along with concepts such as over-fitting, bias-variance trade-off etc. You don't need to extra mathematical background. The explanation is easy-to understand and integrated. If you are a novice on Machine Learning, and non-professional in Machine Learning, this will be a good starting point. Moreover, If you have a good knowledge about ML, or professional, this book will pin-point the important concept that you probably have missed. However, if you need a MATHEMATICALLY RIGOROUS treatment about ML, this book is not. This book is for practitioners who wish to use ML to solve a specific problem. Good companion with this book is "Machine Learning in Action" by Peter Harrington because this book contains no implementation details about the technique but just execution of R. In MLiA there are plenty of implementation code written in Python. Altogether, Highly Recommended!!!

This books teaches the mathematical fundamentals of Machine Learning. It does not cover Neural Networks which are just one implementation (albeit, a very popular one). However, it provides the basis for practitioners to be able to apply various methods to an array of problems. Exercise are in R, which is why I do not give it a 5 star rating. Someone should come up with a Python version of the book.

This book is available for free online, but I found it difficult to read the PDF for long time on computer/tablet, so purchased the book.

The book does a great job of introducing machine learning from a statistical perspective. You also get to learn R. It covers a lot of ground and covers them quite well. Springer has done a wonderful job as usual with their acid-free paper and excellent editors.

CONS: Chapter 3 is somewhat long and the authors keep referring to figures in previous pages. (I almost gave up on the book when reading chapter 3.) It may be handy to keep a decent intro statistics book like Statistics, 4th Edition handy when reading this book. It would have also been nice had the authors referred where some of the basic formulas could be studied in more detail. Another thing is some of the advanced topics could have been moved to separate chapters or appendices (esp the topics the authors skip in their online course). Chapter 7 is quite advanced, uses esoteric techniques and could have been presented after decision trees (ch 8) and SVMs (ch 9).

If you are a Python dev, then go with Python Machine Learning instead. It is more practical, covers most of the practical topics in ISLR, teaches neural nets and some NLP and is presented from a machine learning practitioner's perspective, rather than a statistician's perspective.

If you are a practitioner and use R, you may also want to get Applied Predictive Modeling as a follow-up book. It is written by the author of the caret R machine learning package and discusses more practical issues.

Outstanding. Of all the books that I have used in my predictive analytics courses, this is by far the best. The text and examples are comprehensible and interesting. The smooth integration with R is very helpful. In my view, the best part of the book are the accompanying videos. Hastie and Tibshirani were engaging and entertaining in the video, which helped me to absorb the material. I've tried to think of a good analogy for Hastie and Tibshirani's videos...and the winner is NPR's Car Talk. It was great watching these gurus in the field chatting through the material as if they were out for beers with a friend interested in learning the subject matter. Their humility and humor were engaging. Finally, it is great that they make the pdf of the book, videos, and slides all available for free. Bravo.
Search the web for the free videos.

This book is probably not for experts (Springer has a more advanced book on Statistical Learning) but it's great for people who intermediate intermediate R users with a reasonable grasp of regression. The examples are easy to follow and the explanations are clear. This book is meant as a guide to IMPLEMENTING machine learning techniques in R. It does not cover the theory or the math behind the methods, nor does it offer proofs.

If you need a practical guide to implementing statistical learning in R, this is an excellent choice. Very understandable and accessible.

This book introduces the important topics regarding Classification and Regression in unified fashion along with concepts such as over-fitting, bias-variance trade-off etc. You don't need to extra mathematical background. The explanation is easy-to understand and integrated. If you are a novice on Machine Learning, and non-professional in Machine Learning, this will be a good starting point. Moreover, If you have a good knowledge about ML, or professional, this book will pin-point the important concept that you probably have missed. However, if you need a MATHEMATICALLY RIGOROUS treatment about ML, this book is not. This book is for practitioners who wish to use ML to solve a specific problem. Good companion with this book is "Machine Learning in Action" by Peter Harrington because this book contains no implementation details about the technique but just execution of R. In MLiA there are plenty of implementation code written in Python. Altogether, Highly Recommended!!!

This books teaches the mathematical fundamentals of Machine Learning. It does not cover Neural Networks which are just one implementation (albeit, a very popular one). However, it provides the basis for practitioners to be able to apply various methods to an array of problems. Exercise are in R, which is why I do not give it a 5 star rating. Someone should come up with a Python version of the book.

This book came highly recommended and I now realize why. It's just fantastic. I love the fact that it is the perfect level of detail for me. I also really like the fact that as soon as a topic is introduced, the questions that come to mind are immediately answered!

I can't recommend the kindle edition as I found the typography and rendering to be poor. But the book itself is excellent! Highly recommend this as a great introduction to linear regression, cross validation, classification, and machine learning in general!

This book is probably the best introduction you can get, and you can follow it even with quite a limited math background if needed. You should follow the Stanford free and self-paced course if you need some further help.
In terms of coverage it stays into more "traditional" topics and does not cover neural networks nor deep learning, but I think you should have a reasonable grasp of these before jumping into playing with APIs. I am sure many will disagree :)
So if you want to build a solid understanding of statistical learning I would advise you to get this book.

I was able to easily forecast a range of probabilities using a Monte Carlo simulation to determine the likelihood I would fall asleep in Stats class. What was harder to predict were the clusters of sonombulistic episodes constituting a Poisson distribution. I will likely be changing my major to Art History at semester.

I am a big fan of "Elements of statistical learning", and "Introduction to statistical learning" is true to that book's form, combining effective writing and appealing visual design, and adding the important participatory, get-your-hands-dirty element with R illustrations and exercises.

PS. "Applied predictive modeling" by Kuhn and Johnson is a similar, good book that may be preferred by some readers.

No secret here, this book is great. With a stronger mathematical approach than the two above, it is still accessible if you have been through the basics of calculus, and it describes with ample details the pros and cons of all main statistical models, plus the secret meanings of all these parameters you’re desperately tweaking when you’re not satisfied with your model’s accuracy.

I do wish there was a version for Python though, as I came to dig much deeper the tools Python provides to run these algorithms. Most of the content and of the exercices can nonetheless be adapted on your own to Python though, and it never hurts to remember the R structure and main packages.

More reviews here: [...]