Suche
Lesesoftware
Specials
Info / Kontakt
Deep Learning For Dummies
von: John Paul Mueller, Luca Massaron
For Dummies, 2019
ISBN: 9781119543022 , 371 Seiten
Format: PDF, Online Lesen
Kopierschutz: DRM
Preis: 22,99 EUR
eBook anfordern
Title Page
3
Copyright Page
4
Table of Contents
7
Introduction
15
About This Book
15
Foolish Assumptions
16
Icons Used in This Book
17
Beyond the Book
18
Where to Go from Here
19
Part 1 Discovering Deep Learning
21
Chapter 1 Introducing Deep Learning
23
Defining What Deep Learning Means
24
Starting from Artificial Intelligence
24
Considering the role of AI
26
Focusing on machine learning
29
Moving from machine learning to deep learning
30
Using Deep Learning in the Real World
32
Understanding the concept of learning
32
Performing deep learning tasks
33
Employing deep learning in applications
33
Considering the Deep Learning Programming Environment
33
Overcoming Deep Learning Hype
36
Discovering the start-up ecosystem
36
Knowing when not to use deep learning
36
Chapter 2 Introducing the Machine Learning Principles
39
Defining Machine Learning
40
Understanding how machine learning works
40
Understanding that it’s pure math
41
Learning by different strategies
42
Training, validating, and testing data
44
Looking for generalization
45
Getting to know the limits of bias
46
Keeping model complexity in mind
47
Considering the Many Different Roads to Learning
47
Understanding there is no free lunch
48
Discovering the five main approaches
48
Delving into some different approaches
50
Awaiting the next breakthrough
54
Pondering the True Uses of Machine Learning
54
Understanding machine learning benefits
55
Discovering machine learning limits
57
Chapter 3 Getting and Using Python
59
Working with Python in this Book
60
Obtaining Your Copy of Anaconda
60
Getting Continuum Analytics Anaconda
61
Installing Anaconda on Linux
61
Installing Anaconda on MacOS
62
Installing Anaconda on Windows
63
Downloading the Datasets and Example Code
68
Using Jupyter Notebook
68
Defining the code repository
70
Getting and using datasets
75
Creating the Application
76
Understanding cells
76
Adding documentation cells
77
Using other cell types
78
Understanding the Use of Indentation
79
Adding Comments
80
Understanding comments
81
Using comments to leave yourself reminders
82
Using comments to keep code from executing
83
Getting Help with the Python Language
83
Working in the Cloud
84
Using the Kaggle datasets and kernels
84
Using the Google Colaboratory
84
Chapter 4 Leveraging a Deep Learning Framework
87
Presenting Frameworks
88
Defining the differences
88
Explaining the popularity of frameworks
89
Defining the deep learning framework
91
Choosing a particular framework
92
Working with Low-End Frameworks
93
Caffe2
93
Chainer
94
PyTorch
94
MXNet
95
Microsoft Cognitive Toolkit/CNTK
96
Understanding TensorFlow
96
Grasping why TensorFlow is so good
96
Making TensorFlow easier by using TFLearn
98
Using Keras as the best simplifier
99
Getting your copy of TensorFlow and Keras
100
Fixing the C++ build tools error in Windows
102
Accessing your new environment in Notebook
103
Part 2 Considering Deep Learning Basics
105
Chapter 5 Reviewing Matrix Math and Optimization
107
Revealing the Math You Really Need
108
Working with data
108
Creating and operating with a matrix
109
Understanding Scalar, Vector, and Matrix Operations
110
Creating a matrix
111
Performing matrix multiplication
113
Executing advanced matrix operations
114
Extending analysis to tensors
116
Using vectorization effectively
118
Interpreting Learning as Optimization
119
Exploring cost functions
119
Descending the error curve
120
Learning the right direction
121
Updating
123
Chapter 6 Laying Linear Regression Foundations
125
Combining Variables
126
Working through simple linear regression
126
Advancing to multiple linear regression
127
Including gradient descent
129
Seeing linear regression in action
130
Mixing Variable Types
131
Modeling the responses
131
Modeling the features
132
Dealing with complex relations
133
Switching to Probabilities
135
Specifying a binary response
135
Transforming numeric estimates into probabilities
136
Guessing the Right Features
138
Defining the outcome of incompatible features
138
Solving overfitting using selection and regularization
139
Learning One Example at a Time
141
Using gradient descent
141
Understanding how SGD is different
141
Chapter 7 Introducing Neural Networks
145
Discovering the Incredible Perceptron
146
Understanding perceptron functionality
146
Touching the nonseparability limit
148
Hitting Complexity with Neural Networks
150
Considering the neuron
150
Pushing data with feed-forward
152
Going even deeper into the rabbit hole
154
Using backpropagation to adjust learning
157
Struggling with Overfitting
160
Understanding the problem
160
Opening the black box
160
Chapter 8 Building a Basic Neural Network
163
Understanding Neural Networks
164
Defining the basic architecture
165
Documenting the essential modules
167
Solving a simple problem
169
Looking Under the Hood of Neural Networks
172
Choosing the right activation function
172
Relying on a smart optimizer
174
Setting a working learning rate
175
Chapter 9 Moving to Deep Learning
177
Seeing Data Everywhere
178
Considering the effects of structure
178
Understanding Moore’s implications
179
Considering what Moore’s Law changes
180
Discovering the Benefits of Additional Data
181
Defining the ramifications of data
182
Considering data timeliness and quality
182
Improving Processing Speed
183
Leveraging powerful hardware
184
Making other investments
184
Explaining Deep Learning Differences from Other Forms of AI
185
Adding more layers
186
Changing the activations
188
Adding regularization by dropout
189
Finding Even Smarter Solutions
190
Using online learning
190
Transferring learning
191
Learning end to end
191
Chapter 10 Explaining Convolutional Neural Networks
193
Beginning the CNN Tour with Character Recognition
194
Understanding image basics
194
Explaining How Convolutions Work
197
Understanding convolutions
197
Simplifying the use of pooling
201
Describing the LeNet architecture
202
Detecting Edges and Shapes from Images
207
Visualizing convolutions
208
Unveiling successful architectures
210
Discussing transfer learning
211
Chapter 11 Introducing Recurrent Neural Networks
215
Introducing Recurrent Networks
216
Modeling sequences using memory
216
Recognizing and translating speech
218
Placing the correct caption on pictures
220
Explaining Long Short-Term Memory
221
Defining memory differences
222
Walking through the LSTM architecture
223
Discovering interesting variants
225
Getting the necessary attention
226
Part 3 Interacting with Deep Learning
229
Chapter 12 Performing Image Classification
231
Using Image Classification Challenges
232
Delving into ImageNet and MS COCO
233
Learning the magic of data augmentation
235
Distinguishing Traffic Signs
237
Preparing image data
238
Running a classification task
242
Chapter 13 Learning Advanced CNNs
247
Distinguishing Classification Tasks
248
Performing localization
249
Classifying multiple objects
249
Annotating multiple objects in images
251
Segmenting images
251
Perceiving Objects in Their Surroundings
253
Discovering how RetinaNet works
253
Using the Keras-RetinaNet code
255
Overcoming Adversarial Attacks on Deep Learning Applications
259
Tricking pixels
260
Hacking with stickers and other artifacts
262
Chapter 14 Working on Language Processing
265
Processing Language
266
Defining understanding as tokenization
267
Putting all the documents into a bag
268
Memorizing Sequences that Matter
271
Understanding semantics by word embeddings
271
Using AI for Sentiment Analysis
275
Chapter 15 Generating Music and Visual Art
283
Learning to Imitate Art and Life
284
Transferring an artistic style
285
Reducing the problem to statistics
286
Understanding that deep learning doesn’t create
288
Mimicking an Artist
288
Defining a new piece based on a single artist
288
Combining styles to create new art
290
Visualizing how neural networks dream
290
Using a network to compose music
291
Chapter 16 Building Generative Adversarial Networks
293
Making Networks Compete
294
Finding the key in the competition
294
Achieving more realistic results
296
Considering a Growing Field
303
Inventing realistic pictures of celebrities
303
Enhancing details and image translation
304
Chapter 17 Playing with Deep Reinforcement Learning
307
Playing a Game with Neural Networks
308
Introducing reinforcement learning
308
Simulating game environments
310
Presenting Q-learning
313
Explaining Alpha-Go
316
Determining if you’re going to win
317
Applying self-learning at scale
319
Part 4 The Part of Tens
321
Chapter 18 Ten Applications that Require Deep Learning
323
Restoring Color to Black-and-White Videos and Pictures
324
Approximating Person Poses in Real Time
324
Performing Real-Time Behavior Analysis
325
Translating Languages
326
Estimating Solar Savings Potential
326
Beating People at Computer Games
327
Generating Voices
328
Predicting Demographics
328
Creating Art from Real-World Pictures
329
Forecasting Natural Catastrophes
330
Chapter 19 Ten Must-Have Deep Learning Tools
331
Compiling Math Expressions Using Theano
331
Augmenting TensorFlow Using Keras
332
Dynamically Computing Graphs with Chainer
333
Creating a MATLAB-Like Environment with Torch
333
Performing Tasks Dynamically with PyTorch
334
Accelerating Deep Learning Research Using CUDA
335
Supporting Business Needs with Deeplearning4j
337
Mining Data Using Neural Designer
337
Training Algorithms Using Microsoft Cognitive Toolkit (CNTK)
338
Exploiting Full GPU Capability Using MXNet
339
Chapter 20 Ten Types of Occupations that Use Deep Learning
341
Managing People
341
Improving Medicine
342
Developing New Devices
343
Providing Customer Support
343
Seeing Data in New Ways
344
Performing Analysis Faster
345
Creating a Better Work Environment
345
Researching Obscure or Detailed Information
347
Designing Buildings
347
Enhancing Safety
348
Index
349
EULA
371