Machine Learning:the Art and Science of Algorithms that Make Sense of Data(Chinese Edition)/机器学习

by Peter A. Flach

本书是迄今市面上内容全面的机器学习教材之一，书中汇集了所有用于理解、挖掘和分析数据的先进方法，并且通过数百个精选实例和解说性插图，直观而准确地阐释了这些方法背后的原理，内容涵盖了机器学习的构成要素和机器学习任务、逻辑模型、几何模型、统计模型，以及矩阵分解、ROC分析等时下热点话题。 本书不仅内容丰富，而且图文并茂，无论是新手还是有经验的读者都能从中获益。

Genres: Computer Science, Programming, Nonfiction, Artificial Intelligence, Technology, Technical, Mathematics

280 pages, Paperback

First published September 1, 2012

Reviews

[Gavin]

Short, friendly, smooth, repetitive. First ML book where I didn't feel dumb.

[May Ling]

Interestingly, most books are a mathematical treatment of Machine learning. This is more of a contextual reading with the math baked in. As such is less a mountain of proofs and more context that then shows the equation that matches this.

The earlier chapters likely better if you do not have the Probability, Stats and regression classes. The later chapters might be more challenging for the math impaired. That said, relative to a pure mathematical treatment, it's significantly easier to place the equations with the use case/model technique choice.

[Fuzzball Baggins]

I learned a lot even though it was a bit boring :P

[Matt Chan]

I didn't have computer access for a while, so I was using this book to learn about ML as much as I can. I think the book did what it has set out to do. It was most successful in the chapters where it constrained itself in discussing one type of models at a time. Elsewhere, I think it sacrifice clarify for brevity with some of the mathematics presentation; sometimes I feel like I'm trying to read the author's mind trying to figure out how he got from one line of the algebra to the next (although I suppose this is a common criticism I have for text books). I also feel like it needed a lot of practical examples for each of the models; I had folks asking me about ML stuff I've learned from the book, but I struggle to explain it beyond the way the book has. In order words, I was able to effectively internalize the knowledge yet. Otherwise, I think this is a pretty good entry to people like me who had some stats background while trying to learn more about ML.

[numbworks]

It's not badly written, but it doesn't follow a linear, subsequent way to expose concepts. Every chapter is a collection of notions that are partially related each others, and this style makes the book really difficult to follow and make practical use of it.

[Gary Lang]

There are more "practical" (re)introductions to the field of machine learning but the thing that’s great about this one is that it goes very deep on the math, which you really need to do if you’re going to even think about doing something original in this space. Read other books to run code samples and run cookbook exercises, but read this book to understand the science. 5 stars

[Mohseen]

Its good