"Python neural network programming," the Chinese version of PDF + PDF in English
Links: https://pan.baidu.com/s/1PmRxRv1203iRu-cx5cJu1w extraction code: 8a23
brief introduction · · · · · ·
Neural network is a simulation of the human brain's neural networks, artificial intelligence in order to be able to achieve the kind of machine learning
technology.
This book reveals the concept behind the neural network, and describes how to implement neural network by Python. The book
It is divided into three chapters and two appendices. Chapter 1 introduces the mathematical thought neural networks are used. Chapter 2 describes the
Neural networks realize Python, handwritten digit recognition, and test the performance of the neural network. Chapter 3 led reading
Learn more about those simple neural network, has been observed by the internal trained neural network, to try to further improve
The performance of neural networks, and deepen understanding of the relevant knowledge. Appendix introduced the required calculus
Raspberry Pi and knowledge.
This book is suitable want to engage in research and exploration of the neural network learning reader reference, also suitable for artificial intelligence
Energy, interest and depth learning machine learning and other related fields readers.
About the Author · · · · · ·
He has a bachelor's degree in physics, machine learning and data mining master's degree. He perennial active in the field of technology London, led and organized the London party Python team (nearly 3,000 members).
table of Contents · · · · · ·
Copyright
Copyright Notice
Summary
translator preface
preface
preface
how neural networks work Chapter 1 001
1.1 has a short foot, inch a director 001
1.2 a simple prediction 003
1.3 classifier and predictor is not much difference 008
1.4 Training simple classification 011
1.5 sometimes a classifier is not sufficient to solve the problem 020
1.6 neurons - the nature of computing 024
1.7 033 tracking signal in a neural network
1.8 fairness, matrix multiplication great use 037
1.9 using matrix multiplication the exemplary three-layer neural network 043
1.10 learning heavy weights from a plurality of nodes 051
a plurality of output node backward error propagation 1.11 053
1.12 backpropagation error more layers 054 to
1.13 using reverse error propagation matrix multiplication 058
1.14 our actual 061 how the updated weights
1.15 weight update successful example 077
1.16 ready data 078
Chapter 2 with Python DIY 083
2.1 Python 083
2.2 interactive Python = IPython 084
2.3 gracefully start using Python 085
2.4 using Python to make the neural network 105
2.5 handwritten numbers datasets MNIST 121
Chapter 3 fun-filled 153
3.1 153 own handwriting digital
internal neural networks of the brain 156 3.2
3.3 create new training data: rotate the image 160
3.4 Conclusion 164
Appendix A Introduction to Calculus 165
A.1 a straight line 166
A.2 a slash 168
A.3 a curve 170.
A.4 painted calculus 172
A.5 non-painted calculus 174
A.6 calculus without rendering the chart 177
A.7 patterns 180 [
A.8 function 182 functions
Appendix B using tree berry sent to work 186
B.1 187 installed IPython
B.2 ensure that all work normally 193
B.3 training and testing the neural network 194
B.4 raspberry Pi successful 195