A Low-Cost Ethics Shaping Approach for Designing Reinforcement Learning Agents

  • 2017-12-12 08:35:52
  • Yueh-Hua Wu, Shou-De Lin
  • 8

Abstract

This paper proposes a low-cost, easily realizable strategy to equip areinforcement learning (RL) agent the capability of behaving ethically. Ourmodel allows the designers of RL agents to solely focus on the task to achieve,without having to worry about the implementation of multiple trivial ethicalpatterns to follow. Based on the assumption that the majority of humanbehavior, regardless which goals they are achieving, is ethical, our designintegrates human policy with the RL policy to achieve the target objective withless chance of violating the ethical code that human beings normally obey.

 

Quick Read (beta)

loading the full paper ...