Nurturing promotes the evolution of reinforcement learning in changing environments

Abstract

An agent interacting with its environment may learn to perform complex tasks through reinforcement learning. Reinforcement learning requires exploration of unfamiliar situations, which necessarily involves unknown and potentially dangerous or costly outcomes. Various sorts of external support for the learning agent are possible through investments of time or other resources. Nurturing, one individual investing in the development of another individual with which it has an ongoing relationship, is widely seen in the biological world, often with parents nurturing their offspring. In artificial intelligence, nurturing can be seen as an opportunity to develop both better machine learning algorithms and robots that assist or supervise other robots. Although research into robot-to-robot nurturing is at a very early stage, the hope is that this approach can result in more sophisticated learning systems. The research presented here demonstrates the effectiveness of nurturing through the evolution of the parameters of a reinforcement learning algorithm that is capable of finding good policies in a changing environment.