Chris A. Mattmann - Machine Learning with TensorFlow

Chris A. Mattmann

To comment go to liveBook

Manning

Shelter Island

For more information on this and other Manning titles go to

manning.com

For online information and ordering of these and other Manning books, please visit manning.com. The publisher offers discounts on these books when ordered in quantity.

For more information, please contact

Special Sales Department

Manning Publications Co.

20 Baldwin Road

PO Box 761

Shelter Island, NY 11964

Email: orders@manning.com

2020 by Manning Publications Co. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by means electronic, mechanical, photocopying, or otherwise, without prior written permission of the publisher.

Many of the designations used by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in the book, and Manning Publications was aware of a trademark claim, the designations have been printed in initial caps or all caps.

Recognizing the importance of preserving what has been written, it is Mannings policy to have the books we publish printed on acid-free paper, and we exert our best efforts to that end. Recognizing also our responsibility to conserve the resources of our planet, Manning books are printed on paper that is at least 15 percent recycled and processed without the use of elemental chlorine.

ISBN: 9781617297717

To my wife, Lisa, and to CJ, Heath, and Hailey

An excellent book about TensorFlow, with many examples to get hands-on experience.

Mikal Dautrey, founder and infrastructure consultant, Isitix

A good way to quickly learn TensorFlow!

David Krief, data scientist, Altansia

A great guide to machine learning. It helped launch my third career!

William Wheeler, Java, Scala, Akka and Python developer, TekSystems

Great introduction to machine learning with TensorFlow.

Hamish Dickson, backend developer, DriveTribe

The book helped me jumpstart TensorFlow and work effectively with it.

Ursin Stauss, software developer, Swiss Post

Its been two years since the publication of the first edition of Machine Learning with TensorFlow. Two years is a long time in the field of artificial intelligence.

Today, we are fascinated by a human-language model with more than 80 billion artificial neurons that have learned more than 170 billion parameters. The cost of training such a model is measured in millions of dollars. Lex Fridman of MIT projected that with the improvements in computing and algorithm design, well soon train a model the size of the human brain for less than a few thousand dollars. Just thinkin our near-term future, well train an AI model with the raw capacity of the human brain for less than the cost of a Peloton stationary bike.

Writing a book to capture this fast-moving technology is fraught with risk. By the time Chris wrote a few chapters, researchers likely produced newer, more elegant approaches to solving the same problems. Yet there are perhaps only 10,000 people today who understand AI deeply. You want to jump in, learn, and start using AI in your work. What is one to do?

Buy this bookeven if you have the first edition. Pay special attention to seven new chapters that walk you through fundamental techniques in AI:

• Chapter 6, Sentiment classification: Large movie-review dataset

• Chapter 8, Inferring user activity from Android accelerometer data

• Chapter 10, Part-of-speech tagging and word-sense disambiguation

• Chapter 12, Applying autoencoders: The CIFAR-10 image dataset

• Chapter 15, Building a real-world CNN: VGG -Face and VGG -Face Lite

• Chapter 17, LSTMs and automatic speech recognition

• Chapter 18, Sequence-to-sequence models for chatbots

Chris helps you learn how machines see, hear, speak, write, and feel within our world. He shows how machines can instantly spot that speck of dust on a windshield, much as human eyes do, with autoencoders.

The modeling techniques, which Chris describes with frustratingly delicious, hands-on detail, will persist through time. Theyre fundamental to framing a problem as tensors in, tensors out, flowing through a graph. Framing a problem correctly is far more important than describing the individual details of how it is solved. Expect those details to change and improve rapidly.

Armed with an appreciation of AI modeling, youll be well-prepared to enjoy the rapid, exponential journey forward in artificial intelligence. Welcome to our world! Jump in, have some fun, crank those GPUs, and do your part to assist humanity in solving intelligence. Reimagine our world with smart machinesthen make it so with TensorFlow.

Chris, thanks for taking the time to be our guide, peppered with that godawful dad humor that I just love.

Scott Penberthy, director of Applied AI at Google

Palo Alto

California August 2020

It was about 15 months ago to this day that I sat down with my freshly-minted copy of the first edition of this book, opened it, and dived right in. I currently manage the Artificial Intelligence, Analytics and Innovative Development Division at NASAs Jet Propulsion Laboratory in beautiful Pasadena, California. At the time, however, I was the deputy chief technology officer (CTO) for IT, with a strong background in data science, information retrieval, and software, but only a surface knowledge of the hot topic called machine learning. I had dabbled with it, but never dived deep, as they say. Knowing Manning and its coverage of topics with practicality, in-depth examples, and most of all humor (I desperately seek it in everything ; humor makes things better), I had a good feeling about the book. At the time, it had been almost a full year since Id had the time to read a technical book, let alone sit down and try the code and exercises.

I decided that with this book, I would have to run the code, pull out pencil and paper, and draw matrices, and write things downyou know, learn what I was reading instead of reading but not learning. Whoo, boy, this book was a doozy. It was humorousprobably the easiest introduction to machine learning that I had readand I actually understood it. I remember remarking to my wife one night, This is why all the billionaire CEOs like [Elon] Musk are afraid of AI. I could see its application to a variety of formats, such as text, sound, vision, and speech. And it uses this amazing framework called TensorFlow that I had been hearing so much about.