Edward provides a testbed for rapid experimentation and research with probabilistic models. Here we show how to apply this process for diverse learning tasks.

Bayesian linear regression
A fundamental model for supervised learning.

Batch training
How to train a model using only minibatches of data at a time.

Visualize learning, explore the computational graph, and diagnose training problems.

Automated transformations
Using transformations to easily work over constrained continuous support.

Linear mixed effects models
Linear modeling of fixed and random effects.

Gaussian process classification
Learning a distribution over functions for supervised classification.

Mixture models
Unsupervised learning by clustering data points.

Latent space models
Analyzing connectivity patterns in neural data.

Mixture density networks
A neural density estimator for solving inverse problems.

Generative adversarial networks
Building a deep generative model of MNIST digits.

Probabilistic decoder
A model of latent codes in information theory.

Inference networks
How to amortize computation for training and testing models.

Bayesian neural network
Bayesian analysis with neural networks.

Probabilistic PCA
Dimensionality reduction with latent variables.

If you’re interested in contributing a tutorial, checking out the contributing page.


We also have a community repository for sharing content such as papers, posters, and slides.


For more background and notation, see the pages below.

There are also companion webpages for several papers about Edward.