Improving Generalization for Temporal Difference Learning: The Successor Representation

doi:10.1162/neco.1993.5.4.613

Improving Generalization for Temporal Difference Learning: The Successor Representation

1993-07-01

0

OA

AI

DOI：10.1162/neco.1993.5.4.613

原文链接

查看原文

摘要

En 中文

Estimation of returns over time, the focus of temporal difference (TD) algorithms, imposes particular constraints on good function approximators or representations. Appropriate generalization between states is determined by how similar their successors are, and representations should follow suit. This paper shows how TD machinery can be used to learn such representations, and illustrates, using a navigation task, the appropriately distributed nature of the result.

Improving Generalization for Temporal Difference Learning: The Successor Representation

摘要

AI总结

期刊

机构

引用论文