Stein Variational Gradient Descent

Tensorflow implementation of Stein Variational Gradient Descent (SVGD)

Results

Description

"SVGD is a general purpose variational inference algorithm that forms a natural counterpart of gradient descent for optimization. SVGD iteratively transports a set of particles to match with the target distribution, by applying a form of functional gradient descent that minimizes the KL divergence."

Link

• For official implementation in Matlab and Theano, please visit the authors' project website - SVGD.
• Paper: Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm, NIPS2017 arXiv.