Edward is a Python library for probabilistic modeling, inference, and criticism. It is a testbed for fast experimentation and research with probabilistic models, ranging from classical hierarchical models on small data sets to complex deep probabilistic models on large data sets. Edward fuses three fields: Bayesian statistics and machine learning, deep learning, and probabilistic programming.
It supports modeling with
- Directed graphical models
- Neural networks (via libraries such as Keras and TensorFlow Slim)
- Conditionally specified undirected models
- Bayesian nonparametrics and probabilistic programs
It supports inference with
- Variational inference
- Black box variational inference
- Stochastic variational inference
- Variational auto-encoders
- Inclusive KL divergence: KL(p||q)
- Monte Carlo
- Hamiltonian Monte Carlo
- Stochastic gradient Langevin dynamics
- Metropolis-Hastings
- Marginal optimization (empirical Bayes, marginal maximum likelihood)
- Variational EM
- Maximum a posteriori estimation (penalized maximum likelihood)
- Laplace approximation
It supports criticism of the model and inference with
- Point-based evaluations
- Posterior predictive checks
Edward is built on top of TensorFlow. It enables features such as computational graphs, distributed training, CPU/GPU integration, automatic differentiation, and visualization with TensorBoard.