BriefGPT.xyz
Sep, 2021
可解释的局部树代理策略
Interpretable Local Tree Surrogate Policies
HTML
PDF
John Mern, Sidhart Krishnan, Anil Yildiz, Kyle Hatch, Mykel J. Kochenderfer
TL;DR
本论文提出一种构建可预测的决策树代替神经网络的方法，决策树易于人类理解，能够量化预测未来行为，从而提高决策的可信度和使用范围。
Abstract
high-dimensional policies
, such as those represented by
neural networks
, cannot be reasonably interpreted by humans. This lack of
interpretabilit
→