Universal Value Function Approximators
Tom Schaul, Daniel Horgan, Karol Gregor, David Silver
Abstract
Value functions are a core component of reinforcement learning. The main idea is to to construct a single function approximator V(s; theta) that estimates the long-term reward from any state s, using parameters θ. In this paper we introduce universal value function approximators (UVFAs) V(s,g;theta) that generalise not just over states s but also over goals g. We develop an efficient technique for supervised learning of UVFAs, by factoring observed values into separate embedding vectors for state and goal, and then learning a mapping from s and g to these factored embedding vectors. We show how this technique may be incorporated into a reinforcement learning algorithm that updates the UVFA solely from observed rewards. Finally, we demonstrate that a UVFA can successfully generalise to previously unseen goals.
BibTeX
@InProceedings{pmlr-v37-schaul15,
title = {Universal Value Function Approximators},
author = {Schaul, Tom and Horgan, Daniel and Gregor, Karol and Silver, David},
booktitle = {Proceedings of the 32nd International Conference on Machine Learning},
pages = {1312--1320},
year = {2015},
editor = {Bach, Francis and Blei, David},
volume = {37},
series = {Proceedings of Machine Learning Research},
address = {Lille, France},
month = {07--09 Jul},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v37/schaul15.pdf},
url = {https://proceedings.mlr.press/v37/schaul15.html},
abstract = {Value functions are a core component of reinforcement learning. The main idea is to to construct a single function approximator V(s; theta) that estimates the long-term reward from any state s, using parameters θ. In this paper we introduce universal value function approximators (UVFAs) V(s,g;theta) that generalise not just over states s but also over goals g. We develop an efficient technique for supervised learning of UVFAs, by factoring observed values into separate embedding vectors for state and goal, and then learning a mapping from s and g to these factored embedding vectors. We show how this technique may be incorporated into a reinforcement learning algorithm that updates the UVFA solely from observed rewards. Finally, we demonstrate that a UVFA can successfully generalise to previously unseen goals.}
}