📖
Deep Learning · Theory

Deep Learning

Ian Goodfellow, Yoshua Bengio, Aaron Courville · 2016

🆓 FreeAdvanced★★★★★800 pagesISBN 978-0262035613

The definitive textbook on deep learning theory, co-authored by three pioneers including the inventor of GANs.

Why you should read this
The "Bible" of deep learning. Required reading for anyone doing serious research in neural networks and AI.

Key Topics

Linear AlgebraProbabilityOptimizationFeedforward NetworksRegularizationCNNsRNNsAutoencodersGANsRepresentation Learning

Chapters (10)

1Introduction8 key concepts

AI history, deep learning motivation

2Linear Algebra14 key concepts

Vectors, matrices, eigendecomposition

3Probability & Information Theory12 key concepts

Random variables, distributions, entropy

4Numerical Computation8 key concepts

Overflow, gradient-based optimization

5ML Basics16 key concepts

Capacity, bias-variance, MLE, Bayesian

6Deep Feedforward Networks18 key concepts

Architecture, activation functions, backprop

7Regularization14 key concepts

L1/L2, dropout, data augmentation, early stopping

8Optimization16 key concepts

SGD, Adam, learning rate schedules, batch normalization

9CNNs14 key concepts

Convolution, pooling, architectures

10Sequence Modeling (RNNs)16 key concepts

LSTM, GRU, encoder-decoder, attention

Real-World Applications
  • Self-driving cars
  • Language translation
  • Drug discovery
  • Generative art
Best For
ML researchersPhD studentsAI engineers

Get This Book