Herding Dynamic Weights for Partially Observed Random Field Models
Machine Learning
2012-05-14 v1 Machine Learning
Abstract
Learning the parameters of a (potentially partially observable) random field model is intractable in general. Instead of focussing on a single optimal parameter value we propose to treat parameters as dynamical quantities. We introduce an algorithm to generate complex dynamics for parameters and (both visible and hidden) state vectors. We show that under certain conditions averages computed over trajectories of the proposed dynamical system converge to averages computed over the data. Our "herding dynamics" does not require expensive operations such as exponentiation and is fully deterministic.
Cite
@article{arxiv.1205.2605,
title = {Herding Dynamic Weights for Partially Observed Random Field Models},
author = {Max Welling},
journal= {arXiv preprint arXiv:1205.2605},
year = {2012}
}
Comments
Appears in Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence (UAI2009)