Grokking deep learning

Grokking Deep Learning MEAP V03......Page 1
Copyright......Page 2
Welcome......Page 3
Brief contents......Page 4
Chapter 1: Introducing Deep Learning: Why you should learn it......Page 5
Why should you learn Deep Learning?......Page 6
Why you should read this book!......Page 7
What you need to get started......Page 9
Conclusion and Primer for Chapter 2......Page 10
Chapter 2:
Fundamental Concepts: How do machines learn?......Page 11
What is Deep Learning?......Page 12
What is Machine Learning?......Page 13
Supervised Machine Learning......Page 14
Unsupervised Machine Learning......Page 15
Parametric vs Non-Parametric Learning......Page 16
Supervised Parametric Learning......Page 17
Step 3: Learn the Pattern......Page 18
Unsupervised Parametric Learning......Page 19
Conclusion......Page 20
Chapter 3:
Building Your First Neural Network: Introduction to Gradient Descent......Page 21
What am I going to learn in this chapter?......Page 22
What is a Neural Network?......Page 24
What does a Neural Network do?......Page 25
Does the network make accurate predictions?......Page 26
Why measure error?......Page 27
What the Simplest Form of Neural Network Learning?......Page 28
Characteristics of Hot and Cold Learning......Page 29
Calculating Both direction and amount from error......Page 30
Learning Is Just Reducing Error......Page 32
Let's Back Up And Talk about Functions......Page 34
Relationship Exploration: Hot and Cold......Page 36
A Box With Rods Poking Out of It......Page 37
Derivatives... take Two......Page 38
What you don't really need to know.........Page 39
How to use a derivative to learn......Page 40
Where is our derivative in the code?......Page 41
Learning Method: Gradient Descent......Page 42
Breaking Gradient Descent......Page 43
Divergence......Page 44
Introducing.... Alpha......Page 45
Alpha In Code......Page 46
Memorizing......Page 47
Chapter 4: Neural Networks with Multiple Inputs: Leveraging Matrix Algebra......Page 48
Expanding to Multiple Input Values......Page 49
What We're Going to Build......Page 50
The Street Light Problem......Page 51
Preparing our Data......Page 53
Matrices and the Matrix Relationship......Page 54
Creating a Matrix or Two in Python......Page 57
What is a Neural Network......Page 58
How Does A Neural Network Predict?......Page 59
How Does A Neural Network Learn?......Page 60
Slow down! What is "delta"?......Page 61
Building our Neural Network in Python......Page 62
Teaching Our Neural Network in Python......Page 63
Quick Review from 10,000 Feet......Page 64
Putting it all together in Python (verbose version)......Page 65
Putting it all together in Python (polished version)......Page 66
The Differences Line by Line......Page 67
Learning the whole dataset!......Page 68
Write it from Memory!......Page 69
Chapter 5:
Building Your First "Deep" Neural Network:  Introduction to Backpropagation......Page 70
Neural Networks Learn Correlation......Page 71
Up and Down Pressure......Page 72
Up and Down Pressure (cont.)......Page 73
Edge Case: Overfitting......Page 74
Edge Case: Conflicting Pressure......Page 75
Edge Case: Conflicting Pressure (cont.)......Page 76
Learning Indirect Correlation......Page 77
Creating Our Own Correlation......Page 78
Stacking Neural Networks?......Page 79
Backpropagation: Long Distance Error Attribution......Page 80
Backpropagation: Why does this work?......Page 81
Why The Neural Network Still Doesn't Work......Page 82
The Secret to "Sometimes Correlation"......Page 84
A Quick Break......Page 85
Our New Prediction Code......Page 86
What We Will Build......Page 87
What It Looks Like......Page 88
A Non-Linear Problem......Page 89
A Non-Linear Solution......Page 90