- Paperback: 856 pages
- Publisher: O'Reilly Media; 2 edition (October 15, 2019)
- Language: English
- ISBN-10: 1492032646
- ISBN-13: 978-1492032649
- Product Dimensions: 7 x 1.5 x 9.5 inches
- Shipping Weight: 2.9 pounds (View shipping rates and policies)
- Customer Reviews: 353 customer ratings
- Amazon Best Sellers Rank: #5,203 in Books (See Top 100 in Books)
Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems 2nd Edition
Use the Amazon App to scan ISBNs and compare prices.
Enter your mobile number or email address below and we'll send you a link to download the free Kindle App. Then you can start reading Kindle books on your smartphone, tablet, or computer - no Kindle device required.
To get the free app, enter your mobile phone number.
Customers who viewed this item also viewed these digital items
Customers who bought this item also bought these eBooks
From the Publisher
This book assumes that you have some Python programming experience and that you are familiar with Python’s main scientific libraries, in particular NumPy, Pandas, and Matplotlib.
Also, if you care about what’s under the hood, you should have a reasonable understanding of college-level math as well (calculus, linear algebra, probabilities, and statistics).
More about this book
Machine Learning in Your Projects
So, naturally you are excited about Machine Learning and would love to join the party! Perhaps you would like to give your homemade robot a brain of its own? Make it recognize faces? Or teach it to walk around? Or maybe your company has tons of data (user logs, financial data, production data, machine sensor data, hotline stats, HR reports, etc.), and you could likely unearth some hidden gems if you just knew where to look. With Machine Learning, you could accomplish the following:
- Segment customers and find the best marketing strategy for each group
- Recommend products for each client based on what similar clients bought
- Detect which transactions are likely to be fraudulent
- Forecast next year’s revenue
- And more
Objective and Approach
This book assumes that you know close to nothing about Machine Learning. Its goal is to give you the concepts, tools, and intuition you need to implement programs capable of learning from data. We will cover a large number of techniques, from the simplest and most commonly used (such as linear regression) to some of the Deep Learning techniques that regularly win competitions.
Rather than implementing our own toy versions of each algorithm, we will be using production-ready Python frameworks:
- Scikit-Learn is very easy to use, yet it implements many Machine Learning algorithms efficiently, so it makes for a great entry point to learn Machine Learning.
- TensorFlow is a more complex library for distributed numerical computation. It makes it possible to train & run very large neural networks efficiently by distributing the computations across potentially hundreds of multi-GPU servers. TensorFlow was created at Google and supports many of its large-scale applications. It's been open source since Nov. 2015, with version 2.0 releasing Oct 2019.
- Keras is a high-level Deep Learning API that makes it very simple to train and run neural networks. It can run on top of either TensorFlow, Theano, or Microsoft Cognitive Toolkit (formerly known as CNTK). TensorFlow comes with its own implementation of this API, called tf.keras, which provides support for some advanced TensorFlow features (e.g., the ability to efficiently load data).
|Machine Learning Pocket Reference||Hands-On Unsupervised Learning Using Python||Practical Automated Machine Learning on Azure||Generative Deep Learning|
|Additional Machine Learning From O'Reilly Media||Working with Structured Data in Python||How to Build Applied Machine Learning Solutions from Unlabeled Data||Using Azure Machine Learning to Quickly Build AI Solutions||Teaching Machines to Paint, Write, Compose, and Play|
About the Author
Aurélien Géron is a machine learning consultant and trainer. A former Googler, he led YouTube's video classification team from 2013 to 2016. He was also a founder and CTO of Wifirst (a leading Wireless ISP in France) from 2002 to 2012, and a founder and CTO of two consulting firms -- Polyconseil (telecom, media and strategy) and Kiwisoft (machine learning and data privacy).
There was a problem filtering reviews right now. Please try again later.
EDIT: Just received the print edition of the book and it's in color! The first edition wasn't. This is a pleasant surprise as it makes it easier to read with various charts and graphics.
I was expecting more from the publisher.
If you want to save the planet by ordering an ebook, you will be disappointed with the quality you receive and I believe this is not fair.
Overall I would recommend. It's been much more interesting than I expected.
Whether you are a data scientist looking to start building predictive models in Python, or a software developer looking to become an ML engineer, look no further!
The excellent balance between theory/background and implementation that was present in the first edition is kept, with the essential material additions made (e.g. the unsupervised learning in the "classical ML" part, or the Keras API, which is quickly becoming the most popular way to use TensorFlow).
Needless to say, the Jupyter notes accompanying each chapter are more than helpful.
Also, as a cherry on top, the illustrations in the printed version are now in color, which makes it even easier to read.
In summary, this book is an absolute must-have for a Python-rooted data scientist / ML engineer!
Most important for me, he focuses on explanation over hand-wavy equations that are rampant in other ML books. I say hand-wavy because they typically go like so: "Here's a hard concept. Rather than explain it well, I'll give you some linear algebra and calculus equations, remind you that this is stuff you should have learned in high school, and then move on." Authors probably feel justified in doing this, but after reading a book like this you understand what they are really doing: Skipping the hard-part of breaking difficult concepts down into chunks that can be consumed by a competent programmer, who is perhaps not an expert in "high school" math. Moreover, this author does so without dumbing down the content. That's the mark of someone who well understands both the content and the audience.
This book is long and dense, and serves as both a guide and a reference. It is not a quick read / overview or light reading type book.
Top international reviews
I ordered on Kindle as much prefer reading that way
Recommended if new to ML/DL/NN etc
Im Vergleich zum Vorgänger gibts etwa 250 Seiten mehr (550 alt vs. 800 neu), diese sind etwas dünner (womöglich Probleme mit Textmarkern) dafür aber in Farbe und das Buch als Ganzes fällt somit nicht wesentlich dicker aus.
Besonders die neuen Kapitel über GANs sind für mich ein Highlight, natürlich aber auch die Verwendung von Keras und TensorFlow 2.0.
A única ressalva que faço é também relacionada a sua maior qualidade: o fato do livro ser extremamente prático infelizmente peca na teoria. Dessa forma, recomendo que ele seja utilizado por pessoas que já tenham conhecimento da teoria de ML ou que ele seja utilizado junto a outro material que tenha a teoria mais forte.
Writing a book like this much be an absolutely massive amount of work - the volume of well-organized material (over 800 pages, some 250 pages more than the predecessor), the code, the excellent explanations .. even the use of color is helpful to understanding, not just a decoration.
Complex concepts are explained in digestible pieces, and follow a logical progression. Some sections end in " ..but this has a shortcoming X, which can be fixed by Y .." which is then covered in detail in the next chapter. Where the material becomes too esoteric, pointers are given in the right directions to follow. Many reasearchers can shave months off their projects by studying the book - which can also be read as a reference work - just looking up the material of interest. I started in the middle, went straight into chapter 15 on RNNs, as I need that material now, and immediately felt at home with the discussion without having gone through chapters 0-14. (I'll read those later...)
In addition to compiling executable notebooks with code on Github, the author has helpfully referenced many classical and more recent papers, and collected them on his site. This collection is of great interest in itself. And being so embedded in the ML world as the author is, he can talk with authority on what works and what doesn't. When he says something has been found to work well, on a certain class of problems, one is ready to believe it.
I have been buying computer books for over 40 years, too many, and this book is in my top five, ever. Excellent, excellent book.
O livro foi impresso em cores o que é um diferencial em relação aos livros publicados pela O’Reilly.
I read a sample chapter in PDF and the math typesetting was perfect. Come on Amazon... either have a working format or just use PDFs like the rest of the world...