Contents

• LICENSE
• Deep Learning Tutorials
• Getting Started
  • Datasets
  • Notation
  • A Primer on Supervised Optimization for Deep Learning
  • Theano/Python Tips
• Classifying MNIST digits using Logistic Regression
  • The Model
  • Defining a Loss Function
  • Creating a LogisticRegression class
  • Learning the Model
  • Testing the model
  • Putting it All Together
• Multilayer Perceptron
  • The Model
  • Going from logistic regression to MLP
  • Putting it All Together
  • Tips and Tricks for training MLPs
• Convolutional Neural Networks (LeNet)
  • Motivation
  • Sparse Connectivity
  • Shared Weights
  • Details and Notation
  • The ConvOp
  • MaxPooling
  • The Full Model: LeNet
  • Putting it All Together
  • Running the Code
  • Tips and Tricks
• Denoising Autoencoders (dA)
  • Autoencoders
  • Denoising Autoencoders
  • Putting it All Together
  • Running the Code
• Stacked Denoising Autoencoders (SdA)
  • Stacked Autoencoders
  • Putting it all together
  • Running the Code
  • Tips and Tricks
• Restricted Boltzmann Machines (RBM)
  • Energy-Based Models (EBM)
  • Restricted Boltzmann Machines (RBM)
  • Sampling in an RBM
  • Implementation
  • Results
• Deep Belief Networks
  • Deep Belief Networks
  • Justifying Greedy-Layer Wise Pre-Training
  • Implementation
  • Putting it all together
  • Running the Code
  • Tips and Tricks
• Miscellaneous
  • Plotting Samples and Filters
• References