001 Chapter 1 What is deep learning (28.64 MB)
002 Chapter 1 Learning rules and representations from data (31.8 MB)
003 Chapter 1 Understanding how deep learning works, in three figures (38.64 MB)
004 Chapter 1 Before deep learning A brief history of machine learning (31.1 MB)
005 Chapter 1 Back to neural networks (29.65 MB)
006 Chapter 1 Why deep learning Why now (21.63 MB)
007 Chapter 1 Algorithms (23.69 MB)
008 Chapter 2 The mathematical building blocks of neural networks (21.33 MB)
009 Chapter 2 Data representations for neural networks (18.48 MB)
010 Chapter 2 Real-world examples of data tensors (21.06 MB)
011 Chapter 2 The gears of neural networks Tensor operations (17.78 MB)
012 Chapter 2 Tensor reshaping (12.89 MB)
013 Chapter 2 The engine of neural networks Gradient-based optimization (21.08 MB)
014 Chapter 2 Derivative of a tensor operation The gradient (29.64 MB)
015 Chapter 2 Chaining derivatives The Backpropagation algorithm (22.52 MB)
016 Chapter 2 Looking back at our first example (21.76 MB)
017 Chapter 3 Introduction to Keras and TensorFlow (31.04 MB)
018 Chapter 3 Setting up a deep learning workspace (18.56 MB)
019 Chapter 3 First steps with TensorFlow (29.05 MB)
020 Chapter 3 Anatomy of a neural network Understanding core Keras APIs (23.67 MB)
021 Chapter 3 The "compile" step Configuring the learning process (28.77 MB)
022 Chapter 4 Getting started with neural networks Classification and regression (23.49 MB)
023 Chapter 4 Building your model (26.74 MB)
024 Chapter 4 Classifying newswires A multiclass classification example (22.27 MB)
025 Chapter 4 Predicting house prices A regression example (25.52 MB)
026 Chapter 5 Fundamentals of machine learning (24.74 MB)
027 Chapter 5 The nature of generalization in deep learning (35.72 MB)
028 Chapter 5 Evaluating machine learning models (31.74 MB)
029 Chapter 5 Improving model fit (17.17 MB)
030 Chapter 5 Improving generalization (30.17 MB)
031 Chapter 5 Regularizing your model (27.02 MB)
032 Chapter 6 The universal workflow of machine learning (29.56 MB)
033 Chapter 6 Collect a dataset (39.7 MB)
034 Chapter 6 Develop a model (19.56 MB)
035 Chapter 6 Beat a baseline (17.67 MB)
036 Chapter 6 Deploy the model (34.66 MB)
037 Chapter 6 Monitor your model in the wild (15.34 MB)
038 Chapter 7 Working with Keras A deep dive (28.8 MB)
039 Chapter 7 Subclassing the Model class (14.92 MB)
040 Chapter 7 Using built-in training and evaluation loops (24.87 MB)
041 Chapter 7 Writing your own training and evaluation loops (19.55 MB)
042 Chapter 7 Make it fast with tf function (15.48 MB)
043 Chapter 8 Introduction to deep learning for computer vision (17.86 MB)
044 Chapter 8 The convolution operation (30.83 MB)
045 Chapter 8 Training a convnet from scratch on a small dataset (28.96 MB)
046 Chapter 8 Data preprocessing (25.63 MB)
047 Chapter 8 Leveraging a pretrained model (28.95 MB)
048 Chapter 8 Feature extraction with a pretrained model (27.37 MB)
049 Chapter 9 Advanced deep learning for computer vision (42.23 MB)
050 Chapter 9 Modern convnet architecture patterns (26.84 MB)
051 Chapter 9 Residual connections (24.78 MB)
052 Chapter 9 Depthwise separable convolutions (30.46 MB)
053 Chapter 9 Interpreting what convnets learn (22.91 MB)
054 Chapter 9 Visualizing convnet filters (17.15 MB)
055 Chapter 9 Visualizing heatmaps of class activation (19.59 MB)
056 Chapter 10 Deep learning for timeseries (23.77 MB)
057 Chapter 10 Preparing the data (19.53 MB)
058 Chapter 10 Let's try a basic machine learning model (19.93 MB)
059 Chapter 10 Understanding recurrent neural networks (17.22 MB)
060 Chapter 10 A recurrent layer in Keras (17.59 MB)
061 Chapter 10 Advanced use of recurrent neural networks (25.96 MB)
062 Chapter 10 Using bidirectional RNNs (28.63 MB)
063 Chapter 11 Deep learning for text (25.89 MB)
064 Chapter 11 Preparing text data (19.97 MB)
065 Chapter 11 Vocabulary indexing (21.4 MB)
066 Chapter 11 Two approaches for representing groups of words Sets and sequences (34.45 MB)
067 Chapter 11 Processing words as a sequence The sequence model approach, Part 1 (31.01 MB)
068 Chapter 11 Processing words as a sequence The sequence model approach, Part 2 (23.32 MB)
069 Chapter 11 The Transformer architecture (30.88 MB)
070 Chapter 11 The Transformer encoder (30.75 MB)
071 Chapter 11 Beyond text classification Sequence-to-sequence learning (35.34 MB)
072 Chapter 11 Sequence-to-sequence learning with Transformer (25.81 MB)
073 Chapter 12 Generative deep learning (35.98 MB)
074 Chapter 12 How do you generate sequence data (36.07 MB)
075 Chapter 12 A text-generation callback with variable-temperature sampling (26.5 MB)
076 Chapter 12 DeepDream (25.84 MB)
077 Chapter 12 Neural style transfer (35.66 MB)
078 Chapter 12 Generating images with variational autoencoders (22.42 MB)
079 Chapter 12 Implementing a VAE with Keras (32.76 MB)
080 Chapter 12 A bag of tricks (26 MB)
081 Chapter 13 Best practices for the real world (27.14 MB)
082 Chapter 13 Hyperparameter optimization (33.82 MB)
083 Chapter 13 Scaling-up model training (22.97 MB)
084 Chapter 13 Multi-GPU training (15.65 MB)
085 Chapter 13 TPU training (18.16 MB)
086 Chapter 14 Conclusions (36.42 MB)
087 Chapter 14 Key enabling technologies (28.31 MB)
088 Chapter 14 Key network architectures (26.06 MB)
089 Chapter 14 The limitations of deep learning (27.45 MB)
090 Chapter 14 Local generalization vs extreme generalization (19.3 MB)
091 Chapter 14 The purpose of intelligence (24.22 MB)
092 Chapter 14 Setting the course toward greater generality in AI (31.16 MB)
093 Chapter 14 Implementing intelligence The missing ingredients (28.93 MB)
094 Chapter 14 The missing half of the picture (21.99 MB)
095 Chapter 14 Blending together deep learning and program synthesis (26.9 MB)
096 Chapter 14 Lifelong learning and modular subroutine reuse (38.13 MB)