TensorFlow for Dummies

Title Page......Page 2
Copyright Page......Page 3
Table of Contents......Page 6
About This Book......Page 14
Icons Used in this Book......Page 15
Beyond the Book......Page 16
Where to Go from Here......Page 17
Part 1 Getting to Know TensorFlow......Page 18
Understanding Machine Learning......Page 20
The Development of Machine Learning......Page 21
Statistical regression......Page 22
Reverse engineering the brain......Page 23
Steady progress......Page 24
The rise of big data and deep learning......Page 25
Machine Learning Frameworks......Page 26
Caffe......Page 27
TensorFlow......Page 28
Installing TensorFlow......Page 30
Python and pip/pip3......Page 31
Installing on Mac OS......Page 32
Installing on Windows......Page 33
Exploring the TensorFlow Installation......Page 34
Running Your First Application......Page 35
Launching Hello TensorFlow!......Page 36
Setting the Style......Page 37
Creating Tensors......Page 40
Creating Tensors with Known Values......Page 41
zeros, ones, and fill......Page 43
Creating Tensors with Random Values......Page 44
Transforming Tensors......Page 46
Basic math operations......Page 48
Rounding and comparison......Page 50
Exponents and logarithms......Page 51
Vector and matrix operations......Page 52
Putting Theory into Practice......Page 55
Chapter 4 Executing Graphs in Sessions......Page 58
Forming Graphs......Page 59
Accessing graph data......Page 60
Creating GraphDefs......Page 62
Creating sessions......Page 64
Executing a session......Page 65
Interactive sessions......Page 66
Writing Messages to the Log......Page 67
Visualizing Data with TensorBoard......Page 69
Generating summary data......Page 70
Writing summary data......Page 72
Putting Theory into Practice......Page 75
Chapter 5 Training......Page 78
Formulating the Model......Page 79
Looking at Variables......Page 80
Initializing variables......Page 81
Determining Loss......Page 82
The Optimizer class......Page 83
The GradientDescentOptimizer......Page 84
The MomentumOptimizer......Page 88
The AdagradOptimizer......Page 89
The AdamOptimizer......Page 90
Feeding Data into a Session......Page 91
Defining the feed dictionary......Page 92
Monitoring Steps, Global Steps, and Epochs......Page 93
Saving variables......Page 95
Restoring variables......Page 96
Working with SavedModels......Page 97
Saving a SavedModel......Page 98
Putting Theory into Practice......Page 99
Visualizing the Training Process......Page 102
Session Hooks......Page 103
Creating a session hook......Page 104
Creating a MonitoredSession......Page 106
Putting theory into practice......Page 107
Part 2 Implementing Machine Learning......Page 110
Chapter 6 Analyzing Data with Statistical Regression......Page 112
Linear Regression: Fitting Lines to Data......Page 113
Polynomial Regression: Fitting Polynomials to Data......Page 116
Setting up the problem......Page 118
Defining models with the logistic function......Page 119
Computing loss with maximum likelihood estimation......Page 120
Putting theory into practice......Page 121
The Modified National Institute of Science and Technology (MNIST) Dataset......Page 123
Defining the model with the softmax function......Page 126
Computing loss with cross entropy......Page 127
Putting theory into practice......Page 128
From Neurons to Perceptrons......Page 130
Neurons......Page 131
Perceptrons......Page 132
Weights......Page 134
Bias......Page 135
Activation functions......Page 136
Layers and Deep Learning......Page 140
Layers......Page 141
Training with Backpropagation......Page 142
Implementing Deep Learning......Page 144
Tuning the Neural Network......Page 146
Input standardization......Page 147
Weight initialization......Page 148
Batch normalization......Page 149
Regularization......Page 152
Variable scope......Page 154
Retrieving variables from collections......Page 155
Improving the Deep Learning Process......Page 156
Creating tuned layers......Page 157
Putting theory into practice......Page 158
Filtering Images......Page 162
Convolution......Page 163
Averaging Filter......Page 164
Filters and features......Page 165
Setting convolution parameters......Page 166
Convolutional Neural Networks (CNNs)......Page 168
Creating convolution layers......Page 169
Creating pooling layers......Page 171
Processing CIFAR images......Page 173
Classifying CIFAR images in code......Page 175
Converting images......Page 179
Color processing......Page 182
Rotating and mirroring......Page 183
Resizing and cropping......Page 185
Convolution......Page 187
Putting Theory into Practice......Page 188
Chapter 9 Analyzing Sequential Data with Recurrent Neural Networks (RNNs)......Page 192
Recurrent Neural Networks (RNNs)......Page 193
RNNs and recursive functions......Page 194
Training RNNs......Page 195
Creating RNN Cells......Page 196
Creating a basic RNN......Page 198
Predicting text with RNNs......Page 201
Creating multilayered cells......Page 203
Creating dynamic RNNs......Page 204
Long Short-Term Memory (LSTM) Cells......Page 205
Creating LSTMs in code......Page 207
Gated Recurrent Units (GRUs)......Page 209
Creating GRUs in code......Page 210
Predicting text with GRUs......Page 211
Part 3 Simplifying and Accelerating TensorFlow......Page 212
Datasets......Page 214
Creating datasets......Page 215
Processing datasets......Page 221
One-shot iterators......Page 226
Initializable iterators......Page 228
Reinitializable iterators......Page 229
Feedable iterators......Page 230
Putting Theory into Practice......Page 231
Bizarro Datasets......Page 234
Loading data from CSV files......Page 235
Loading the Iris and Boston datasets......Page 236
Chapter 11 Using Threads, Devices, and Clusters......Page 238
Configuring a new session......Page 239
Configuring a running session......Page 241
Building TensorFlow from source......Page 242
Assigning operations to devices......Page 248
Configuring GPU usage......Page 250
Executing TensorFlow in a Cluster......Page 251
Creating a ClusterSpec......Page 252
Creating a server......Page 253
Specifying jobs and tasks......Page 254
Running a simple cluster......Page 257
Chapter 12 Developing Applications with Estimators......Page 260
Training an Estimator......Page 261
Running an Estimator......Page 263
Creating Input Functions......Page 264
Configuring an Estimator......Page 265
Using Feature Columns......Page 266
Creating and Using Estimators......Page 269
Linear regressors......Page 270
DNN classifiers......Page 273
Combined linear-DNN classifiers......Page 275
Wide and deep learning......Page 276
Analyzing census data......Page 277
Running Estimators in a Cluster......Page 282
Accessing Experiments......Page 283
Creating an experiment......Page 284
Methods of the experiment class......Page 285
Running an experiment......Page 286
Putting theory into practice......Page 287
Chapter 13 Running Applications on the Google Cloud Platform (GCP)......Page 290
Working with GCP projects......Page 291
Billing......Page 292
The Cloud Software Development Kit (SDK)......Page 293
The gcloud Utility......Page 294
Buckets......Page 296
Objects and virtual hierarchy......Page 298
The gsutil utility......Page 299
Receiving arguments......Page 303
Packaging TensorFlow code......Page 304
Executing Applications with the Cloud SDK......Page 306
Local execution......Page 307
Deploying to the cloud......Page 308
Configuring a Cluster in the Cloud......Page 312
Setting the training input......Page 313
Obtaining the training output......Page 316
Setting the prediction input......Page 317
Obtaining the prediction output......Page 318
Part 4 The Part of Tens......Page 320
Tensor......Page 322
Graph......Page 323
Variable......Page 324
Dataset......Page 325
Saver......Page 326
Select a Representative Dataset......Page 328
Start with a Small Number of Layers......Page 329
Normalize Batch Data......Page 330
Check Weights and Gradients......Page 331
Index......Page 332
EULA......Page 0