Planning with Arithmetic and Geometric Attributes
Artificial Intelligence
2018-09-07 v1
Abstract
A desirable property of an intelligent agent is its ability to understand its environment to quickly generalize to novel tasks and compose simpler tasks into more complex ones. If the environment has geometric or arithmetic structure, the agent should exploit these for faster generalization. Building on recent work that augments the environment with user-specified attributes, we show that further equipping these attributes with the appropriate geometric and arithmetic structure brings substantial gains in sample complexity.
Cite
@article{arxiv.1809.02031,
title = {Planning with Arithmetic and Geometric Attributes},
author = {David Folqué and Sainbayar Sukhbaatar and Arthur Szlam and Joan Bruna},
journal= {arXiv preprint arXiv:1809.02031},
year = {2018}
}