Contents
Preface xiii
Acknowledgements xv
1 Introduction 1
2 Supervised learning 17
3 Shallow neural networks 25
4 Deep neural networks 41
5 Loss functions 56
6 Fitting models 77
7 Gradients and initialization 96
8 Measuring performance 118
9 Regularization 138
10 Convolutional networks 161
11 Residual networks 186
12 Transformers 207
13 Graph neural networks 240
14 Unsupervised learning 268
15 Generative Adversarial Networks 275
16 Normalizing flows 303
17 Variational autoencoders 326
18 Diffusion models 348
19 Reinforcement learning 373
20 Why does deep learning work? 401
21 Deep learning and ethics 420
A Notation 436
B Mathematics 439
C Probability 448
Bibliography 462
Index 513