John Mueller has produced hundreds of books and articles on topics ranging from networking to home security and from database management to heads-down programming.

Luca Massaron is a senior expert in data science who has been involved with quantitative methods since 2000. He is a Google Developer Expert (GDE) in machine learning.

Introduction 1

About This Book 1

Foolish Assumptions 2

Icons Used in This Book 3

Beyond the Book 3

Where to Go from Here 4

Part 1: Introducing How Machines Learn 5

Chapter 1: Getting the Real Story about AI 7

Moving beyond the Hype 8

Dreaming of Electric Sheep 9

Understanding the history of AI and machine learning 10

Exploring what machine learning can do for AI 11

Considering the goals of machine learning 12

Defining machine learning limits based on hardware 12

Overcoming AI Fantasies 13

Discovering the fad uses of AI and machine learning 14

Considering the true uses of AI and machine learning 15

Being useful; being mundane 16

Considering the Relationship between AI and Machine Learning 17

Considering AI and Machine Learning Specifications 18

Defining the Divide between Art and Engineering 19

Predicting the Next AI Winter 20

Chapter 2: Learning in the Age of Big Data 23

Considering the Machine Learning Essentials 24

Defining Big Data 25

Considering the Sources of Big Data 26

Building a new data source 26

Using existing data sources 29

Locating test data sources 29

Specifying the Role of Statistics in Machine Learning 30

Understanding the Role of Algorithms 31

Defining what algorithms do 32

Considering the five main techniques 32

Defining What Training Means 34

Chapter 3: Having a Glance at the Future 37

Creating Useful Technologies for the Future 38

Considering the role of machine learning in robots 38

Using machine learning in health care 39

Creating smart systems for various needs 40

Using machine learning in industrial settings 40

Understanding the role of updated processors and other hardware 41

Discovering the New Work Opportunities with Machine Learning 42

Working for a machine 42

Working with machines 43

Repairing machines 44

Creating new machine learning tasks 44

Devising new machine learning environments 45

Avoiding the Potential Pitfalls of Future Technologies 46

Part 2: Preparing Your Learning Tools 47

Chapter 4: Installing a Python Distribution 49

Using Anaconda for Machine Learning 50

Getting Anaconda 50

Defining why Anaconda is used in this book 51

Installing Anaconda on Linux 52

Installing Anaconda on Mac OS X 53

Installing Anaconda on Windows 54

Downloading the Datasets and Example Code 57

Using Jupyter Notebook 57

Defining the code repository 59

Understanding the datasets used in this book 64

Chapter 5: Beyond Basic Coding in Python 67

Defining the Basics You Should Know 68

Considering Python basics 68

Working with functions 72

Working with modules 76

Storing Data Using Sets, Lists, and Tuples 78

Creating sets 78

Performing operations on sets 78

Using lists 79

Creating and using tuples 82

Defining Useful Iterators 83

Working with ranges 83

Iterating multiple lists using zip 84

Working with generators using yield 84

Indexing Data Using Dictionaries 85

Creating dictionaries 85

Storing and retrieving data from dictionaries 85

Chapter 6: Working with Google Colab 87

Defining Google Colab 88

Understanding what Google Colab does 88

Considering the online coding difference 90

Using local runtime support 91

Working with Google Colab features 91

Getting a Google Account 94

Creating the account 94

Signing in 95

Working with Notebooks 96

Creating a new notebook 96

Opening existing notebooks 97

Uploading a notebook 99

Saving notebooks 100

Downloading notebooks 103

Performing Common Tasks 103

Creating code cells 104

Creating text cells 106

Creating special cells 107

Editing cells 108

Moving cells 108

Using Hardware Acceleration 108

Viewing Your Notebook 109

Displaying the table of contents 110

Getting notebook information 110

Checking code execution 110

Executing the Code 111

Sharing Your Notebook 112

Getting Help 113

Part 3: Getting Started with the Math Basics 115

Chapter 7: Demystifying the Math Behind Machine Learning 117

Working with Data 118

Learning the terminology 119

Understanding scalar and vector operations 120

Performing vector multiplication 123

Creating a matrix 123

Understanding basic operations 125

Performing matrix multiplication 126

Glancing at advanced matrix operations 128

Using vectorization effectively 129

Exploring the World of Probabilities 130

Getting an overview of probability 130

Operating on probabilities 131

Conditioning chance by Bayes’ theorem 132

Describing the Use of Statistics 135

Chapter 8: Descending the Gradient 139

Acknowledging Different Kinds of Learning 140

Supervised learning 140

Unsupervised learning 141

Reinforcement learning 141

The learning process 142

Mapping an unknown function 142

Exploring cost functions 145

Descending the optimization curve 147

Optimizing with big data 148

Leveraging sampling 149

Using parallelism 150

Learning out-of-core 151

Chapter 9: Validating Machine Learning 153

Considering the Use of Example Data 154

Checking Out-of-Sample Errors 155

Understanding the concept of samples 155

Looking for the holy grail of generalization 156

Experimenting how bias and variance work 158

Keeping model complexity in mind 160

Keeping solutions balanced 162

Depicting learning curves 163

Training, Validating, and Testing 165

Considering the split 165

Resorting to cross-validation 166

Looking for alternatives in validation 167

Optimizing by Cross-Validation 169

Sources of predictive performance 169

Exploring the hyper-parameter space 170

Selecting relevant features 171

Avoiding Sample Bias and Leakage Traps 173

Chapter 10: Starting with Simple Learners 175

Discovering the Incredible Perceptron 176

Falling short of a miracle 176

Hitting the nonseparability limit 179

Growing Greedy Classification Trees 180

Predicting outcomes by splitting data 181

Pruning overgrown trees 185

Taking a Probabilistic Turn 188

Understanding Naïve Bayes 189

Estimating response with Naïve Bayes 192

Part 4: Learning from Smart and Big Data 197

Chapter 11: Preprocessing Data 199

Gathering and Cleaning Data 200

Repairing Missing Data 201

Identifying missing data 201

Choosing the right replacement strategy 203

Transforming Distributions 205

Creating Your Own Features 207

Understanding the need to create features 207

Creating features automatically 208

Explaining the basics of SVD 210

Reorganizing data 212

Delimiting Anomalous Data 215

Using a univariate strategy                .215

Resorting to Multivariate Models 217

Chapter 12: Leveraging Similarity 221

Measuring Similarity between Vectors 222

Understanding similarity 222

Computing distances for learning 223

Using Distances to Locate Clusters 224

Checking assumptions and expectations 226

Inspecting the gears of the K-means algorithm 227

Tuning the K-Means Algorithm 229

Experimenting with K-means reliability 230

Experimenting with how centroids converge 233

Finding Similarity by K-Nearest Neighbors 238

Understanding the k parameter 238

Experimenting with a flexible algorithm 240

Chapter 13: Working with Linear Models the Easy Way 243

Starting to Combine Features 244

Getting an overview of regression 244

Solving problems with a machine learning approach 247

Understanding R squared 249

Mixing Features of Different Types 251

Switching to Probabilities 255

Specifying a binary response 255

Handling multiple classes 259

Guessing the Right Features 259

Defining the outcome of features that don’t work together 259

Solving overfitting by using greedy selection 260

Addressing overfitting by regularization 262

Learning One Example at a Time 264

Using gradient descent 264

Understanding how SGD is different 265

Chapter 14: Hitting Complexity with Neural Networks 271

Revising the Perceptron 272

Pushing forth with feed-forward 274

Going even deeper down the rabbit hole 276

Pulling back with backpropagation 280

Representing the Way of Learning of a Network 283

Understanding the problem with overfitting 283

Choosing a framework 285

Getting your copy of TensorFlow and Keras 286

Opening the black box 289

Introducing Deep Learning 294

Understanding some deep learning essentials 295

Explaining the magic of convolutions 296

Understanding recurrent neural networks 300

Chapter 15: Going a Step Beyond Using Support Vector Machines 307

Revisiting the Separation Problem 308

Explaining the Algorithm 309

Avoiding the pitfalls of nonseparability 311

Applying nonlinearity 312

Explaining the kernel trick by example 314

Classifying and Estimating with SVM 316

Chapter 16: Resorting to Ensembles of Learners 319

Leveraging Decision Trees 320

Growing a forest of trees 322

Understanding the importance measures 327

Working with Almost Random Guesses 330

Bagging predictors with Adaboost 330

Boosting Smart Predictors 333

Meeting again with gradient descent 334

Considering the state of the art in tabular data 335

Averaging Different Predictors 336

Blending solutions 337

Stacking diverse solutions 337

Part 5: Applying Learning to Real Problems 339

Chapter 17: Classifying Images 341

Working with a Set of Images 342

Revising the State of the Art in Computer Vision 347

Extracting Visual Features 350

Recognizing Faces Using Eigenfaces 352

Classifying Images 356

Chapter 18: Scoring Opinions and Sentiments 361

Introducing Natural Language Processing 362

Revising the State of the Art in NLP 363

Understanding How Machines Read 364

Defining the input data 364

Processing and enhancing text 366

Scraping textual datasets from the web 371

Handling problems with raw text 374

Using Scoring and Classification 375

Performing classification tasks 375

Analyzing reviews from e-commerce 378

Chapter 19: Recommending Products and Movies 383

Realizing the Revolution of E-Commerce 384

Downloading Rating Data 386

Trudging through the MovieLens dataset 386

Navigating through anonymous web data 390

Encountering the limits of rating data 392

Considering collaborative filtering 392

Catching the Limits of Behavioral Data 397

Integrating Text and Behaviors 399

Viewing the attributes 399

Obtaining statistics 400

Leveraging SVD 400

Understanding the SVD connection 400

Seeing SVD in action 401

Part 6: The Part of Tens 405

Chapter 20: Ten Ways to Improve Your Machine Learning Models 407

Studying Learning Curves 408

Using Cross-Validation Correctly 409

Choosing the Right Error or Score Metric 410

Searching for the Best Hyper-Parameters 410

Testing Multiple Models 411

Averaging Models 411

Stacking Models 412

Applying Feature Engineering 412

Selecting Features and Examples 413

Looking for More Data 414

Chapter 21: Ten Guidelines for Ethical Data Usage 415

Obtaining Permission 416

Using Sanitization Techniques 417

Avoiding Data Inference 418

Using Generalizations Correctly 418

Shunning Discriminatory Practices 419

Detecting Black Swans in Code 420

Understanding the Process 420

Considering the Consequences of an Action 421

Balancing Decision Making 421

Verifying a Data Source 422

Chapter 22: Ten Machine Learning Packages to Master 423

Gensim 423

imbalanced-learn 424

OpenCV 424

SciPy 425

SHAP 426

Statsmodels 427

Modin 427

PyTorch 428

Poetry 429

Snorkel 429

Index 431

The New copy of this book will include any supplemental materials advertised. Please check the title of the book to determine if it should include any access cards, study guides, lab manuals, CDs, etc.

The Used, Rental and eBook copies of this book are not guaranteed to include any supplemental materials. Typically, only the book itself is included. This is true even if the title states it includes any access cards, study guides, lab manuals, CDs, etc.