Variational Deep Q Network
Machine Learning
2017-12-01 v1 Artificial Intelligence
Machine Learning
Abstract
We propose a framework that directly tackles the probability distribution of the value function parameters in Deep Q Network (DQN), with powerful variational inference subroutines to approximate the posterior of the parameters. We will establish the equivalence between our proposed surrogate objective and variational inference loss. Our new algorithm achieves efficient exploration and performs well on large scale chain Markov Decision Process (MDP).
Cite
@article{arxiv.1711.11225,
title = {Variational Deep Q Network},
author = {Yunhao Tang and Alp Kucukelbir},
journal= {arXiv preprint arXiv:1711.11225},
year = {2017}
}
Comments
12 pages, 5 figures, Second workshop on Bayesian Deep Learning (NIPS 2017)