0%

We're in the midst of an AI research explosion. Deep learning has unlocked superhuman perception that has powered our push toward self-driving vehicles, the ability to defeat human experts at a variety of difficult games including Go and Starcraft, and even generate essays with shockingly coherent prose. But deciphering these breakthroughs often takes a Ph.D. education in machine learning and mathematics.

This updated second edition describes the intuition behind these innovations without the jargon and complexity. By the end of this book, Python-proficient programmers, software engineering professionals, and computer science majors will be able to re-implement these breakthroughs on their own and reason about them with a level of sophistication that rivals some of the best in the field.

New chapters cover recent advancements in the fields of generative modeling and interpretability. Code examples throughout the book are updated to TensorFlow 2 and PyTorch 1.4.

Table of Contents

  1. 1. Fundamentals of Linear Algebra
    1. Data Structures and Operations 
    2. The Fundamental Spaces
    3. Eigenvectors and Eigenvalues
    4. Summary
  2. 2. Fundamentals of Probability
    1. Events and Probability
    2. Conditional Probability
    3. Random Variables
    4. Expectation and Variance
    5. Bayes’ Theorem
    6. Entropy, Cross-Entropy, and KL Divergence
    7. Continuous Probability Distributions
    8. Summary
  3. 3. Training Feed-Forward Neural Networks
    1. The Fast-Food Problem
    2. Gradient Descent
    3. The Delta Rule and Learning Rates
    4. Gradient Descent with Sigmoidal Neurons
    5. The Backpropagation Algorithm
    6. Stochastic and Minibatch Gradient Descent
    7. Test Sets, Validation Sets, and Overfitting
    8. Preventing Overfitting in Deep Neural Networks
    9. (Optional) Theory
    10. Summary
  4. 4. Convolutional Neural Networks
    1. Neurons in Human Vision
    2. The Shortcomings of Feature Selection
    3. Vanilla Deep Neural Networks Don’t Scale
    4. Filters and Feature Maps
    5. Full Description of the Convolutional Layer
    6. Max Pooling
    7. Full Architectural Description of Convolution Networks
    8. Closing the Loop on MNIST with Convolutional Networks
    9. Image Preprocessing Pipelines Enable More Robust Models
    10. Accelerating Training with Batch Normalization
    11. Building a Convolutional Network for CIFAR-10
    12. Visualizing Learning in Convolutional Networks
    13. Residual learning and skip connections for very deep networks
    14. Leveraging Convolutional Filters to Replicate Artistic Styles
    15. Learning Convolutional Filters for Other Problem Domains
    16. Summary
  5. 5. Models for Sequence Analysis
    1. Analyzing Variable-Length Inputs
    2. Tackling seq2seq with Neural N-Grams
    3. Implementing a Part-of-Speech Tagger
    4. Dependency Parsing and SyntaxNet
    5. Beam Search and Global Normalization
    6. A Case for Stateful Deep Learning Models
    7. Recurrent Neural Networks
    8. The Challenges with Vanishing Gradients
    9. Long Short-Term Memory (LSTM) Units
    10. TensorFlow Primitives for RNN Models
    11. Implementing a Sentiment Analysis Model
    12. Solving seq2seq Tasks with Recurrent Neural Networks
    13. Augmenting Recurrent Networks with Attention
    14. Dissecting a Neural Translation Network
    15. Self-attention and Transformers
    16. Summary
  6. 6. Generative Models
    1. Generative Adversarial Networks
    2. Variational Autoencoders
    3. Implementing a VAE
    4. Score-based Generative Models
    5. Denoising Autoencoders and Score Matching
    6. Summary
  7. 7. Methods in Interpretability
    1. Overview
    2. Summary of classical interpretable models
    3. Methods for evaluating feature importance
    4. Extractive Rationalization
    5. LIME
    6. SHAP
    7. Summary
  8. 8. Deep Reinforcement Learning
    1. Deep Reinforcement Learning Masters Atari Games
    2. What Is Reinforcement Learning?
    3. Markov Decision Processes (MDP)
    4. Policy
    5. Future Return
    6. Discounted Future Return
    7. Explore Versus Exploit
    8. Policy Versus Value Learning
    9. Policy Learning via Policy Gradients
    10. Pole-Cart with Policy Gradients
    11. OpenAI Gym
    12. Creating an Agent
    13. Building the Model and Optimizer
    14. Sampling Actions
    15. Keeping Track of History
    16. Policy Gradient Main Function
    17. PGAgent Performance on Pole-Cart
    18. Trust-region Policy Optimization
    19. Proximal Policy Optimization
    20. Q-Learning and Deep Q-Networks
    21. The Bellman Equation
    22. Issues with Value Iteration
    23. Approximating the Q-Function
    24. Deep Q-Network (DQN)
    25. Training DQN
    26. Learning Stability
    27. Target Q-Network
    28. Experience Replay
    29. From Q-Function to Policy
    30. DQN and the Markov Assumption
    31. DQN’s Solution to the Markov Assumption
    32. Playing Breakout wth DQN
    33. Building Our Architecture
    34. Stacking Frames
    35. Setting Up Training Operations
    36. Updating Our Target Q-Network
    37. Implementing Experience Replay
    38. DQN Main Loop
    39. DQNAgent Results on Breakout
    40. Improving and Moving Beyond DQN
    41. Deep Recurrent Q-Networks (DRQN)
    42. Asynchronous Advantage Actor-Critic Agent (A3C)
    43. UNsupervised REinforcement and Auxiliary Learning (UNREAL)
    44. Summary
  9. Index
3.238.142.134