Efficient Methods for Unsupervised Learning of Probabilistic Models

05/19/2012

∙

by   Jascha Sohl-Dickstein, et al.

∙

0

∙

share

In this thesis I develop a variety of techniques to train, evaluate, and sample from intractable and high dimensional probabilistic models. Abstract exceeds arXiv space limitations -- see PDF.