Introduction to Machine Learning

The topics and concepts in this book are exceptionally well organized. After reading it from cover to cover, I could easily see how all the ideas and concepts fit into place. I have two main criticisms. First, the notation is sometimes non-standard, e.g. the r vector is used to denote the label vector and superscripts are used sometimes as subscripts. Second, the explanations are sometimes too brief. For example, when deriving the solution for Least Squares Regression with Quadratic Discriminants, Vandermode matrices are used but the author fails to identify them as such, or to explain why they are useful. If the author were to write an extra sentence on every other page, the explanations would be perfect!

I have a little knowledge about some areas of Machine Learning; I have found this book to be a very useful reference for the areas that I am not familiar with. Explanations are very clear with a very nice examples and illustrations; author also provides good references if deeper understanding of the topic is desired. Each chapter has a notes section which I found particularly useful, since it gives a brief overview of the field with good references. Author nicely ties all of the topics together so a more deeper and wholesome understanding could be obtained. I would highly recommend this book to both undergraduate and graduate students who are interested in Machine Learning. P.S. I am a PhD candidate in Computer Science.

I was very happy with this book. The author used good judgement when deciding the level of detail to delve into for each concept. I was not brand new to machine learning but I still got alot out of the book.

I would like to congratulate the author on writing this book, which is crisp and covers whole range of topics. What I liked the most is a systematic disucssion on a wide variety of areas in machine learning with a certain degree of details. But at the same time, I will also say that the book at some places,(for eg the treatment of Multi Dimensional scaling and Linear discriminants analysis,) lacks depth in its derivations. Also if some explanatory examples are put,it would help the reader, who is doing a first time reading, in understanding the concepts. At the same time, I think the book achieves it's target of introducing to the reader, a whole gamet of techniques, at a fairly reasonable level. The book is no doubt, a nice and one-stop quick reference for many topics, as such. A commendable thing is an up to date errata maintained by the author, with latest editions made. I would recommend the book for a quick introduction to the subject.

We are only beginning to teach silicon based computers how to do things that meat computers have been doing for many thousands of years, things like talking. We learn to talk by making mistakes. Babies gurgle and cry and once in a while make a word. Momma, Daddy and Grandma reward the baby and eventually he's chattering away. Silicon brains can't do that. But with the advances in computer technology, we are gaining the ability to store and process large amounts of data, as well as to access it from physically distant locations. With this mass of data, we have made progress in "data mining." If a person buys the first Harry Potter book. there's a percentage that will buy the second, and a different percentage that will buy the third. You can mine the data for these numbers. And by analyzing these percentages you can determine the likelihood of success in directing advertising to this customer. This is just one example of machine learning. Other topics covered in this book include statistics, pattern recognition, neural networks, artificial intelligence, signal processing, process control. This book is intended for the beginning student in machine learning, he should have some background in programming, probability, calculus, and linear algebra. Having said that, I can recommend this book to anyone moving into the machine learning area.