从摘要数据中进行反向强化学习

Mar, 2017

Inverse Reinforcement Learning from Summary Data

Antti Kangasrääsiö, Samuel Kaski

TL;DR该论文介绍了一种新的逆向强化学习方法，该方法不需要对数据进行特定的简化假设，可以在不确定简化函数的情况下进行推理和评估参数不确定性。

Abstract

inverse reinforcement learning (IRL) aims to explain observed complex behavior by fitting reinforcement learning models to behavioral data. However, traditional IRL methods are only applicable when the observatio