Uncertainty and Exploration

doi:10.1101/265504

Uncertainty and Exploration

2018-02-14

0

OA

AI

DOI：10.1101/265504

原文链接

原文求助

摘要

En 中文

AbstractIn order to discover the most rewarding actions, agents must collect information about their environment, potentially foregoing reward. The optimal solution to this “explore-exploit” dilemma is often computationally challenging, but principled algorithmic approximations exist. These approximations utilize uncertainty about action values in different ways. Somerandomexploration algorithms scale the level of choice stochasticity with the level of uncertainty. Otherdirectedexploration algorithms add a “bonus” to action values with high uncertainty. Random exploration algorithms are sensitive tototaluncertainty across actions, whereas directed exploration algorithms are sensitive torelativeuncertainty. This paper reports a multi-armed bandit experiment in which total and relative uncertainty were orthogonally manipulated. We found that humans employ both exploration strategies, and that these strategies are independently controlled by different uncertainty computations.

Uncertainty and Exploration

摘要

期刊

机构

引用论文