Module: tf.contrib.bayesflow

Defined in tensorflow/contrib/bayesflow/

Ops for representing Bayesian computation.

This package provides classes for Bayesian computation with TensorFlow.


custom_grad module: Functions for specifying custom gradients.

hmc module: Hamiltonian Monte Carlo, a gradient-based MCMC algorithm.

metropolis_hastings module: Functions to create a Markov Chain Monte Carlo Metropolis step.

monte_carlo module: Monte Carlo integration and helpers.

optimizers module: Probabilistic optimizer modules.