TensorFlow 1.x Deep Learning Cookbook: Over 90 unique recipes to solve artificial-intelligence driven problems with Python

Excellent book for a high level overview in moderate to semi complex things in TF. That way it's not so overwhelming to get hands-on on that level. (Beyond beginner hello world stuff).

Good product good read

Wishing to learn about TensorFlow, I decided to survey TF books available from Amazon, and pick one or two for further study. I excluded self-published offerings, and ended up with this longish list, dominated by Packt titles:"Machine Learning with TensorFlow" by Shukla, published by Manning in 2018-02, 272 pp, $43"Mastering TensorFlow 1.x" by Fandango, Packt, 2018-01, 474 pp, $35"Pro Deep Learning with TensorFlow" by Pattanayak, Apress, 2017-12, 398 pp, $37"TensorFlow 1.x Deep Learning Cookbook" by Gulli and Kapoor, Packt, 2017-12, 536 pp, $32"Neural Network Programming with TensorFlow" by Ghotra and Dua, Packt, 2017-11, 274 pp, $40"Predictive Analytics with TensorFlow" by Karim, Packt, 2017-11, 522 pp, $50"Machine Learning with TensorFlow 1.x" by Hua and Azeem, Packt, 2017-11, 304 pp, $39"Learning TensorFlow" by Hope and Resheff, O'Reilly, 2017-08, 242 pp, $25"Hands-On Deep Learning with TensorFlow" by Van Boxel, Packt, 2017-07, 174 pp, $35"Deep Learning with TensorFlow" by Zaccone, Karim, Menshawy, Packt, 2017-04, 320 pp, $50"TensorFlow Machine Learning Cookbook" by McClure, Packt, 2017-02, 370 pp, $30"Building Machine Learning Projects with TensorFlow" by Bonnin, Packt, 2016-11, 291 pp, $35"Getting Started with TensorFlow" by Zaccone, Packt, 2016-07, 180 pp, $35I reviewed the doc on tensorflow.org - including the doc for older releases - then started looking at books.Two weeks later, I am still not done - the book by Shukla has not arrived - but the picture is reasonably clear. The books by Zaccone, Karim, Zaccone and Karim, Bonnin, Hua and Azeem, Ghotra and Dua, and (probably) Van Boxel, can be skipped. (See my reviews of those titles for detail). The remaining five choices fall into four clusters. First, the book by Hope and Resheff provides a good-quality introduction to TensorFlow. Second, the book by Pattanayak unexpectedly goes for academic rigor - the book's subtitle refers to "mathematical foundations" - and emerges as a textbook about the algorithms associated with TensorFlow. The third cluster is formed by the books by Fandango and Gulli and Kapoor - both are unpolished but serviceable, substantial books which go for wide coverage. Finally, McClure's book sits between Clusters 1 and 3.If I want to continue this elimination game, "Fandango vs. Gulli and Kapoor" is an obvious match-up - and Fandango comes out on top, although only on points. Fandango's book is thinner, but covers more topics - if I had to put a number on it, I would say it covers 10-15% more. G&K's seem to have a markedly superior coverage of CNNs - but that topic is just not my cup of tea. Writing is very similar across the two books. In the end, I pick Fandango and move on, but you may sensibly decide otherwise.

I am posting this as an anonymous reviewer because I am downright embarrassed to have bought this book, even direct from the publisher for only five dollars. But the editors, technical editors, copy editors and so forth should be even more embarrassed to charge money for this-- it is so riddled with editing errors as to be useless. And these are errors of every magnitude and type:There are trivial errors in the code section of the book, where variable names are mangled through elementary cut and paste errors or bad formatting.There are formatting errors in the code sections of the book. This is TensorFlow, therefore Python, and whitespace and indenting in Python matters. But that doesn't stop the authors and typesetters from interspersing prose comments in between indented sections of code in a way that utterly confuses the whitespace.There are serious errors in the code section of the book, where even once you figure out what the variable name is supposed to be, still fail to execute because the overall concept is Just Plain Wrong.There are errors in the prose part of the book, some easily understood, some painfully awkward, and some which will cause the reader to stop and puzzle out exactly what is being said.There are organization errors in the prose where, for instance, the authors say something like, "There are three ways to do such-and-such," then give poorly explained examples in a numbered list, then barrel right on in that numbered list for totally unrelated points four, five and six. No, really-- it's a marvel to behold. I promise you, no one read this book between delivery of the raw manuscript and final publication. Not one person. Just like no one actually ran the code in this book.Once we get to meaningful examples, the examples are woefully incomplete. Some-- not all-- of the code is available on the author's private GitHub page, but not all of it, and major pieces of exposition are missing. How does one actually obtain a particular CSV file used in the example, for instance? Well, you can get it from the GitHub page, but this is nowhere explained. Or maybe they keep calling it a CSV file but you're supposed to use the raw, non-CSV raw file that they link to from another site entirely. Who can know? Not you, not from reading this book!Even the typesetting is atrocious-- in addition to the indenting/whitespace issues above, there are occasional snippets of mathematical notation, and in the PDF, at least, they look like someone took very low resolution bitmap graphics and jammed them into the text looking pixelated and sad. This is what we've come to: A book so terrible that the deficiencies of the typesetting are relevant.This book is trash and morally speaking these huckster publishers owe me more than my original five dollar investment back as payment for my stolen time.

Without using complicated maths, authors have very meticulously implemented deep neural networks.The book is not for novices; instead, it is for advanced users. You will make most of the book if you:1. Are advanced Python users and looking to build an understanding of Deep Neural Networks. OR2. Know deep learning, and want to shift to TensorFlow for your work.The recipes use standard datasets already built into TensorFlow library, and this allows one to focus on implementation details of the model.I especially liked the chapter on Autoencoders as it very exhaustively covered the implementation of different types of Autoencoders on the same dataset allowing me to compare their performance. The chapter on Reinforcement learning was refreshingly simple, I could copy the recipe and with little modification had my policy gradient model working in the different game environment in no time.The book also gives step by step instruction on how to use Tensorflow models for mobile apps both on IOS and Android.The USP of the book is that it covers almost all DL models starting from the basic MLP to the recently released CAPSNet. Another great thing about the book is that it has given the references to the research papers it has implemented. The hands-on approach of the book makes it a useful reference for research.For codes, I would suggest you use the GitHub link provided in the book.Overall it is a great book.

I bought this book to expand my knowledge of tensorflow. In general this is good collage of articles on neural networks. However, all the articles separately available openly in the internet. For example chapter on lstms there is google neural translation machine, I expected to get more insight than it is in original google article, but the paragraph in the book is just copy/paste.There are plenty of typos in text as well as in code.Overall, if you are just starting with tensorflow, it’s a good reference point, however It is overpriced for the content that is freely available.

Covers a range of ideas, but is scrappily put together, with entire paragraphs duplicated in several chapters (e.g. Explanation of transfer learning). Proofreading of code is laughably bad, in many places the code has got mixed up with the non-code, fonts mixed up, code line endings are missing, etc. There's cheaper ways of learning the same stuff.Conclusion : maybe try the "Hands on" book by Gulon instead. I bought both and learnt more from that book.

This is highly recommended book. But Amazon shipped me India edition while charged for international edition. Instead of published 35% discount I ended up paying 35% more on printed price. Since I am in urgent need of the book, I am not going to return it.

Il est inadmissible de vendre presque au prix neuf un kindle illisible !