Welcome to breze’s documentation!

Basics

• Specifiying losses, norms, transfer functions etc.
  • Using the builtin loss functions
  • Using custom loss functions
  • Using norms and transfer functions

Models and Algorithms

Learning representations, clustering:

• Principal Component Analysis
• Extreme Component Analysis
• Sparse Filtering
• ICA with Reconstruction Cost
• Canonical Correlation Analysis
• Slow Feature Analysis
• K-Means
• Regularized Information Maximization
• Stochastic Gradient Variational Bayes

Denoising:

• Linear Denoiser

Supervised Learning

• Recurrent Neural Networks
• Multilayer Perceptrons

Sampling

• Hybrid Monte Carlo

Trainers

• Trainers
  • Trainer module
  • Score module
  • Report module

Helpers, convenience functions and tools

• Feature extraction
• Data manipulation
• Various utilities
• Helpers for plotting data

Architectures, Components

• Norms
• Transfer functions
• Loss functions
• Stochastic Corruption of Theano Variables
• Miscellaneous functionality
• Layers
• Common functions
• Univariate Normal Distribution
• Multivariate Normal Distribution
• Utilities
  • Nested Lists for Theano, etc.
  • GPU related utilities
  • Other

For variance propagation:

• Common functions

Implementation Notes

• Variance propagation
  • References
  • Recurrent Networks
  • Transfer functions
  • Losses

Indices and tables

• Index
• Module Index
• Search Page