Winter Driving Best Books of the Month Valentine's Day Shop Learn more nav_sap_SWP_6M_fly_beacon King easycohice_2016 All-New Amazon Fire TV Subscribe & Save Valentine's Day Cards Amazon Gift Card Offer chiraq chiraq chiraq  Amazon Echo All-New Fire Kindle Paperwhite Prime Exclusive Savings in Video Games Shop Now Sale

Customer Reviews

4.5 out of 5 stars44
Format: PaperbackChange
Price:$49.49+ Free shipping with Amazon Prime
Your rating(Clear)Rate this item


There was a problem filtering reviews right now. Please try again later.

on January 30, 2014
Some of the other reviewers seem to be unclear on the concept of "introduction", demanding rigor and depth - and, unfortunately, not suggesting any alternatives. If you want deep understanding of the algorithms, you will need to start with a proper textbook, like "Pattern matching" by Bishop, or "Elements of statistical learning" by Hastie and Tibshirani. If you want a more accessible, high-level presentation with examples that can be reproduced, in R, you want to get "Introduction to statistical learning" by James, Witten, Hastie and Tibshirani. If you want a *really* accessible introduction to the techniques - again, with examples that you can try and build on - well, "Machine learning with R" just might be the best choice today. (On R side; if you have invested in Python, Peter Harrington's "Machine learning in action" and books by Wes McKinney are a good bet). OK, this is not an outstanding book, it is under-edited and plain-looking - unfortunately, Packt follow the no-frills approach of O'Reilly - but it is friendly, reasonably well written, and offers a good deal of content. (Extra brownie points for Chapter 11). Let me put it this way: you want to read "Introduction to statistical learning", but "Machine learning with R" is a good warm-up.
33 comments88 of 91 people found this helpful. Was this review helpful to you?YesNoReport abuse
on April 6, 2014
First off, I am newbie to both machine learning and R and wanted find a starting point somewhere. I browsed around many books before deciding on this one. The writing style of Mr. Lantz is provided in a very understandable/readable manner. It's akin to someone sitting next to you and explaining things in a down to earth, layman's fashion rather than try to "tech speak" you to death with complicated explanations (aka formal textbook). Just the right amount of hand holding for me. I highlight quite bit and it's actually difficult with this book as there isn't much fluff. He's very succinct. The books states that it's for someone who know some ML and no R or R and no ML. I don't know either and the material is digestible except for one thing: review your stats! I took statistics long ago in college and never really learned it well the first time so I had stop and reread core concepts before continuing. Do yourself a favor and review basic statistics and probability before you start this book. I read both "Naked Statistics" and "Statistics in Plain English" and it helped me a great deal (and probably will continue to do so since it appears a bulk of machine learning is stats and prob). Currently into about a third of the way in and I am finding it to be very enjoyable and practical. Other reviewers point out that this book is too basic and this may be the case, but for someone like me who is starting from absolute scratch and who needs to understand basic ML concepts (AND basic R) I find it a great book. Will post an addendum once I complete it.
0Comment34 of 35 people found this helpful. Was this review helpful to you?YesNoReport abuse
on June 15, 2014
This book uses R packages that are have been updated since its publication and no longer work with the code given in the book. I contacted the publisher, but because the code works fine with the package versions it was written for, they will not offer updates on their website. If you know machine learning and R well, you can probably figure out a workaround, but you're also not the intended audience for this book.
44 comments47 of 53 people found this helpful. Was this review helpful to you?YesNoReport abuse
on February 13, 2014
If you are new to both machine learning and R and want to learn both at the same time, I can't imagine there being a better book.

I needed to figure out how to implement nearest neighbors, decision trees, SVM, neural networks, and boosting on two data sets, in a short amount of time. I had no experience with R and my only prior experience with machine learning was neural networks. Using this book I was able to implement four algorithms in R. For each topic the book describes an application, the algorithm, provides code to implement the algorithm. You can download the data set from the publisher's website so you can try it out.
0Comment15 of 17 people found this helpful. Was this review helpful to you?YesNoReport abuse
on March 20, 2014
This is a great book. I liked the way authors highlight syntax for models and discuss strengths and weeknesses. It has a nice balance of theory and hands-on training. However, I would need to use a R book, such as R in action, in conjunction with it.

I have looked at many books on the topic. I will put my review for all of these. Perhaps this can save you some time.
1) http://www.amazon.com/dp/0470650931 : Good theoretical book, but badly written and does not have any hands on exercise.
2) http://www.amazon.com/dp/1466503963 : This is another great book. Good balance of theory and hands-on exercise. This is an excellent book to start learning data mining and R. However, this book relies on a GUI RCommander. It does a good job and one can do a lot with it but it has its limitations. However, I will still use this book.
3) http://www.amazon.com/dp/1439810184 : This is an advanced book and heavily entrinched in cases. This makes it difficult to replicate things unless your work is directly related to one of the case studies covered.
4) http://www.amazon.com/dp/0133412938 : Good examples, but does not explain much about the interpretation. This leaves one wondering what is the purpose of certain graph, what are the axis and how to interpret it. if appropriate explanation is added, this would be an excellent book.
5) http://www.amazon.com/dp/111844714X : This book is very expensive and almost totally devoid of any theory or discussion. I would not use it.
6) http://www.amazon.com/dp/1441998896: This is a decent book. It relies on another GUI, Rattle. It is a strong contender to the book 2 in this list.
22 comments26 of 33 people found this helpful. Was this review helpful to you?YesNoReport abuse
on May 8, 2014
This book is fantastic. I've been through the first four chapters, so I can't give a complete review yet, but if the rest of the book conains surprises, I will update.

Chapter 3 is on k-nearest neighbor and chapter 4 is on naive Bayes. Both provide a clear explanation of the technique, and its pros and cons, along with a simple, non-programming example. Then there is an excellent walkthrough of applying the technique to real world data, and some hints at how to improve the results by adjusting the inputs. After each chapter, I felt that I was ready to apply the technique, with the understanding that I did not yet understand everything about the topic. The chapters also are careful to explain that 1) the particular R package used has more options than have been covered and 2) there are other R packages to be investigated and experimented with.

I just couldn't be happier with these chapters, and I am fired up to keep reading and working along with the examples. It has been years since I have been reading a tech book and thinking, "OK, just a little bit more, I don't want to quit yet" instead of, "Have I spent enough time and effort on this for one day yet?" It's that good.

Although the book does provide a bit of an introduction to R, it is by no means comprehensive, and I have found that R takes time, patience and frustration to master (not that I've mastered it yet). So if I had been a complete R newbie, I probably would have struggled more.

I also found that a recent course I took in statistical inference was helpful, although not really required. At one point the book describes normalizing features using Z scores. I knew what Z scores were, so that was a point I didn't need to simply take on faith.

If you are relatively new to R and machine learning, and you want to learn to do machine learning with R, this *is* the book.

UPDATE: I am working through chapter 10, on validating model performance, and it is a goldmine. Maybe it's a matter of my general knowledge about machine learning concepts coalescing at the same time I'm reading this chapter, but it was like a long line of lightbulbs were getting switched on in my head.

UPDATE 2: I've now read the entire book and chapter 11, on improving model performance, was invaluable. It introduces random forests, and somehow finds a way to make the concept easy to comprehend, where every other article I've read or video I've watched made it seem very complicated. This chapter has given me the confidence to start tackling more sophisticated kaggle competitions using ensemble methods. Really looking forward to it.
0Comment1 of 1 people found this helpful. Was this review helpful to you?YesNoReport abuse
on March 13, 2014
The writing is sparse and clear. The examples and the code are straightforward and easy to understand. I wish I had read this years ago when I was first learning to use R. If you are new to R, definitely buy this book.
0Comment2 of 2 people found this helpful. Was this review helpful to you?YesNoReport abuse
on November 16, 2014
I have a large collection of books on programming, R, and machine learning and I am constantly looking for new material on state of the art practices related to data science. I think this is one of the best in terms of readability, straightforward and practical examples that demonstrate the key concepts in real-world terms, and up-to-date information about the use of advanced R packages for parallel processing and very large datasets. Unlike many other books on the subject, author Brett Lantz presents the material in a crisp and clear manner and does an excellent job integrating the machine learning concepts, underlying statistical foundations, R programming constructs, and practical examples. I think this will be a constant reference in my work. I look forward to future publications from this author and hope he has a blog or other means to keep up with his ideas and insights.
0Comment1 of 1 people found this helpful. Was this review helpful to you?YesNoReport abuse
on December 14, 2014
I really like this book. It provides nice introduction to machine learning and stat methods. Provides theoretical background, introduces interesting problems, and shows you how to use R in your daily job. It is good that author provides references to the raw data and useful tips.
Book helped me personally to learn new data mining methods. The only thing which I miss is some sort of training material, like exercises or control questions.
Bottom line - useful book for good price. I recommend it.
0Comment1 of 1 people found this helpful. Was this review helpful to you?YesNoReport abuse
on September 28, 2014
The best part of this book is that it not only introduces how to model a machine learning problem for a dataset, explains the individual lines of code, but also goes ahead and explains how you can improve the model for higher accuracy - that is something unique about the approach in this book.
0Comment1 of 1 people found this helpful. Was this review helpful to you?YesNoReport abuse

Send us feedback

How can we make Amazon Customer Reviews better for you?
Let us know here.