Efficient Methods for Unsupervised Learning of Probabilistic Models
Machine Learning
2012-05-22 v1 Artificial Intelligence
Information Theory
Neural and Evolutionary Computing
math.IT
Data Analysis, Statistics and Probability
Abstract
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.
Cite
@article{arxiv.1205.4295,
title = {Efficient Methods for Unsupervised Learning of Probabilistic Models},
author = {Jascha Sohl-Dickstein},
journal= {arXiv preprint arXiv:1205.4295},
year = {2012}
}