Machine Learning: The Art and Science of Algorithms that Make Sense of Data 1st Edition

What an amazing book, I got it about a month ago for a self-study routine and every page of this book has been a joy. I am an undergraduate CS major with a decent amount of math experience, and for me this book is a tough but rewarding read. I constantly find myself reading the same section 2 or 3 times in a row, restling with the concepts until I can grasp some intuition of the topics bring discussed. The author is very thorough in their writing, making sure to fill in the details so you dont get left behind in the mathematical notation. The book is filled with beautiful graphs and other figures to further help the reader along in their understanding of machine learning.As a heads up, this book is heavy on the theory and light on the application, so keep that in mind when considering this book for purchase. It isn't going to give you a simple recipe to plug into R. It did however, lay out the intricacies of machine learning in a very abstract and methodical fashion, allowing the reader to gain a much deeper insight into the mechanics of the popular ML techniques than a more practical book would.

I have purchased 5 books on Machine Learning - and this is the best one. Of course you need some mathematical background, but this book is highly readable and explains concepts in a great way

In real world, three cohorts would approach Machine Learning differently -A. Programmers - "How" - interested in quickly learning the libraries, tips/tricks to scale algorithms with larger data setsB. Theorists - "What" - interested in choosing the right algorithm, design ensemble, selecting and extracting right featuresC. Fashionists - "Show" - in this category, some of the even basic reporting/analytics are not termed "Machine Learning", need enough buzzwords pieced together to repaint the old apps.Flach's book is a great source for those who are 75%-25% between first two, and perhaps even greater especially if your Linear Algebra (basics) is not too rusty. It gives a wide and somewhat deep tour of the landscape broken into four paradigms (Quantitative/Analytical, Logical, Geometric, Probabilitisic) and does a real good job on feature design. The book is interspersed with some key insights that are not to be found elsewhere (e.g., how the 'pseudo-inverse' in OLS is really decorrelate-scale-normalize the distribution; Skew-Kurtosis are the statistical measure of "shape"; Naive Bayes is not only Naive but also not particularly Bayesian; How Laplacian Estimate generalizes into Pseudo-Counts and then to m-estimate etc.). After "deep reading" of the book over a month or so, I also went through Flach's detailed 500+ slide presentation (check out his website) on this book. It was very useful to improve solutions several key machine learning problems at work. Flach especially shines on usage of ROC to algorithm comparison which has been his key research area.A few items that I think would've nailed 5-stars -1. Total omission of Neural Nets (ANNs)2. Only a glimpse of RBF while discussing the generalization from kNN to GMMs - as a key activation function more detailed treatment on RBF would help.3. Flach does a really good job of summarizing - at the end of each chapter and at the end of the book - the key insights. A similar "Real World Insights", which are interspersed in the book (e.g., how Naive Bayes is a GREAT classifier, but lousy estimator), aggregated would have helped.Overall, going back in time, I would buy and study it again. For a great first book, I recommend Hastie's "An Introduction to Statistical Learning", or Hal Duame's "A Course In Machine Learning" (ciml.info). After finishing this book, I would recommend "Pattern Classification" (Duda, Stork) which further elaborates on most stuff here and also has a great elucidation on Neural Networks.

At first. I am rewriting review for this book. This books covers fundamental theory about ML. So I was very frustracted because lack of mathmetical background. But once you get use to it. This book will be definite guide book for ML study. Admitting..this book is hard to read. but worth it. and can't be easier because ML itself is VERY hard topic !!!

This looks like a very nice text, but the figures are badly done: in particular, items on them (scatter points, labels, ticks) are unacceptably small. What probably happened (speaking from experience of having done similar things) was that the author made larger plot which were than shrunk by the editor to fit on the page. The size of the labels and datapoints should've been adjusted to anticipate for the shrinkage.

Clearly written. Shows the theory and practice of machine learning. An invaluable tutorial.

The subtitle "the art and science of algorithms that make sense of data" is completely misleading and the main reason I am rating it two stars.A more accurate subtitle would be "mathematical foundations of machine learning".There is little in the way of algorithms and when present they are very high level algorithms.The book is fairly well written, but not suitable as a first book of machine learning.Buy this book only if you have grasped the intuition of how the basic machine learning algorithm work and want to go deeper into their mathematical foundations .Avoid this book if :you want an overview of ML algorithmsyou want to explor ML use casesyou want to explore the workflow of ML proyects.

This book is an excellent introduction to machine learning complete with the math necessary to understand the subject area. I highly recommend it.