Artificial Intelligence with Python Cookbook: Proven recipes for applying AI algorithms and deep learning techniques using TensorFlow 2.x and PyTorch 1.6

459 Pages·2020·14.479 MB·English

by  Ben Auffarth

Artificial Intelligence with
Python Cookbook
Proven recipes for applying AI algorithms and deep learning
techniques using TensorFlow 2.x and PyTorch 1.6
Ben Auffarth
BIRMINGHAM - MUMBAI
Artificial Intelligence with Python Cookbook
Copyright © 2020 Packt Publishing
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form
or by any means, without the prior written permission of the publisher, except in the case of brief quotations
embedded in critical articles or reviews.
Every effort has been made in the preparation of this book to ensure the accuracy of the information presented.
However, the information contained in this book is sold without warranty, either express or implied. Neither the
author, nor Packt Publishing or its dealers and distributors, will be held liable for any damages caused or alleged to
have been caused directly or indirectly by this book.
Packt Publishing has endeavored to provide trademark information about all of the companies and products
mentioned in this book by the appropriate use of capitals. However, Packt Publishing cannot guarantee the accuracy
of this information.
Commissioning Editor: Sunith Shetty
Acquisition Editor: Siddharth Mandal
Content Development Editor: Nathanya Dias
Senior Editor: David Sugarman
Technical Editor: Sonam Pandey
Copy Editor: Safis Editing
Project Coordinator: Aishwarya Mohan
Proofreader: Safis Editing
Indexer: Priyanka Dhadke
Production Designer: Jyoti Chauhan
First published: October 2020
Production reference: 1291020
Published by Packt Publishing Ltd.
Livery Place
35 Livery Street
Birmingham
B3 2PB, UK.
ISBN 978-1-78913-396-7
www.packt.com
Packt.com
Subscribe to our online digital library for full access to over 7,000 books and videos, as well
as industry leading tools to help you plan your personal development and advance your
career. For more information, please visit our website.
Why subscribe?
Spend less time learning and more time coding with practical eBooks and videos
from over 4,000 industry professionals
Improve your learning with skill plans tailored especially for you
Get a free eBook or video every month
Fully searchable for easy access to vital information
Copy and paste, print, and bookmark content
Did you know that Packt offers eBook versions of every book published, with PDF and
ePub files available? You can upgrade to the eBook version at www.packt.com and, as a
print book customer, you are entitled to a discount on the eBook copy. Get in touch with us
at [email protected] for more details.
At www.packt.com, you can also read a collection of free technical articles, sign up for a
range of free newsletters, and receive exclusive discounts and offers on Packt books and
eBooks.
Contributors
About the author
Ben Auffarth is a full-stack data scientist with more than 15 years of work experience. With
a background and Ph.D. in computational and cognitive neuroscience, he has designed and
conducted wet lab experiments on cell cultures, analyzed experiments with terabytes of
data, run brain models on IBM supercomputers with up to 64k cores, built production
systems processing hundreds of thousands of transactions per day, and trained neural
networks on millions of text documents. He resides in West London with his family, where
you might find him in a playground with his young son. He co-founded and is the former
president of Data Science Speakers, London.
I am deeply grateful to the editors at Packt, who provided practical help and competent
advice, and to everyone who has been close to me and supported me, especially my partner
Diane. This book is dedicated to Diane and my son, Nicholas.
About the reviewers
va barbosa is a software engineer in the Qiskit Community at IBM, focused on building
open source tools and creating educational content for developers, researchers, students,
and educators in the field of quantum computing. Previously, va was a developer advocate
at the Center for Open-Source Data and AI Technologies where he assisted developers
discover and make use of data science and machine learning technologies. He is fueled by
his passion to help others and guided by his enthusiasm for open source technology.
Eyal Wirsansky is an Artificial Intelligence consultant and mentor, a senior software
engineer, and a technology community leader. Eyal leads the Jacksonville (FL) Java User
Group, hosts the Artificial Intelligence for Enterprise virtual user group, writes a
developer-oriented Artificial Intelligence blog, and often speaks about topics related to
Artificial Intelligence, Machine Learning and Genetic Algorithms. Eyal is the author of the
book Hands-On Genetic Algorithms with Python.
Packt is searching for authors like you
If you're interested in becoming an author for Packt, please visit authors.packtpub.com
and apply today. We have worked with thousands of developers and tech professionals,
just like you, to help them share their insight with the global tech community. You can
make a general application, apply for a specific hot topic that we are recruiting an author
for, or submit your own idea.
Table of Contents
Preface 1
Chapter 1: Getting Started with Artificial Intelligence in Python 8
Technical requirements 8
Setting up a Jupyter environment 9
Getting ready 9
How to do it... 10
Installing libraries with Google Colab 10
Self-hosting a Jupyter Notebook environment 11
How it works... 14
There's more... 14
See also 15
Getting proficient in Python for AI 16
Getting ready 17
How to do it... 18
Obtaining the history of Jupyter commands and outputs 18
Execution history 19
Outputs 19
Auto-reloading packages 19
Debugging 20
Timing code execution 21
Displaying progress bars 22
Compiling your code 23
Speeding up pandas DataFrames 25
Parallelizing your code 26
See also 28
Classifying in scikit-learn, Keras, and PyTorch 29
Getting ready 29
How to do it... 30
Visualizing data in seaborn 31
Modeling in scikit-learn 33
Modeling in Keras 35
Modeling in PyTorch 42
How it works... 45
Neural network training 45
The SELU activation function 49
Softmax activation 49
Cross-entropy 49
See also 50
Modeling with Keras 51
Getting ready 52
How to do it... 53
Table of Contents
Data loading and preprocessing 54
Model training 62
How it works... 66
Maximal information coefficient 66
Data generators 67
Permutation importance 68
See also 68
Chapter 2: Advanced Topics in Supervised Machine Learning 70
Technical requirements 71
Transforming data in scikit-learn 71
Getting ready 72
How to do it... 73
Encoding ranges numerically 73
Deriving higher-order features 76
Combining transformations 78
How it works... 80
There's more... 81
See also 82
Predicting house prices in PyTorch 82
Getting ready 82
How to do it... 85
How it works... 92
There's more... 93
See also 95
Live decisioning customer values 96
Getting ready 96
How to do it... 97
How it works... 100
Active learning 101
Hoeffding Tree 101
Class weighting 101
See also 102
Battling algorithmic bias 102
Getting ready 103
How to do it... 105
How it works... 113
There's more... 114
See also 117
Forecasting CO2 time series 118
Getting ready 118
How to do it... 119
Analyzing time series using ARIMA and SARIMA 121
How it works... 123
There's more... 126
See also 127
[ ii ]
Table of Contents
Chapter 3: Patterns, Outliers, and Recommendations 129
Clustering market segments 130
Getting ready 130
How to do it... 131
How it works... 136
There's more... 138
See also 140
Discovering anomalies 141
Getting ready 141
How to do it... 142
How it works... 149
k-nearest neighbors 149
Isolation forest 149
Autoencoder 149
See also 150
Representing for similarity search 150
Getting ready 151
How to do it... 152
Baseline – string comparison functions 153
Bag-of-characters approach 154
Siamese neural network approach 155
How it works... 159
Recommending products 160
Getting ready 160
How to do it... 162
How it works... 164
Precision at k 165
Matrix factorization 165
The lightfm model 166
See also 166
Spotting fraudster communities 167
Getting ready 167
How to do it... 168
Creating an adjacency matrix 168
Community detection algorithms 169
Evaluating the communities 170
How it works... 172
Graph community algorithms 172
Louvain algorithm 172
Girvan–Newman algorithm 173
Information entropy 173
There's more... 173
See also 175
Chapter 4: Probabilistic Modeling 176
Technical requirements 177
Predicting stock prices with confidence 177
[ iii ]
Table of Contents
Getting ready 177
How to do it... 178
How it works... 182
Featurization 182
Platt scaling 183
Isotonic regression 184
Naive Bayes 184
See also 185
Estimating customer lifetime value 186
Getting ready 186
How to do it... 186
How it works... 188
The BG/NBD model 189
The Gamma-Gamma model 189
See also 189
Diagnosing a disease 190
Getting ready 190
How to do it... 191
How it works... 196
Aleatoric uncertainty 196
Negative log-likelihood 197
Bernoulli distribution 197
Metrics 197
See also 198
Stopping credit defaults 198
Getting ready 198
How to do it... 199
How it works... 203
Epistemic uncertainty 203
See also 204
Chapter 5: Heuristic Search Techniques and Logical Inference 205
Making decisions based on knowledge 205
Getting ready 206
How to do it... 206
Logical reasoning 207
Knowledge embedding 207
How it works... 211
Logical reasoning 212
Logic provers 213
Knowledge embedding 213
Graph embedding with Walklets 213
See also 214
Solving the n-queens problem 216
Getting ready 216
How to do it... 217
Genetic algorithm 217
[ iv ]