Policy Fusion for Adaptive and Customizable Reinforcement Learning Agents
This work addresses the need for customizable and adaptive agents in video game production, offering practical methods to reduce reliance on manually designed reward functions, though it is incremental in applying existing techniques to this domain.
The paper tackles the problem of training reinforcement learning agents for game development that produce meaningful interactions with players and exhibit designer-specified behavioral traits, by proposing policy fusion methods to combine pre-trained policies, with entropy-weighted fusion outperforming others in experiments.
In this article we study the problem of training intelligent agents using Reinforcement Learning for the purpose of game development. Unlike systems built to replace human players and to achieve super-human performance, our agents aim to produce meaningful interactions with the player, and at the same time demonstrate behavioral traits as desired by game designers. We show how to combine distinct behavioral policies to obtain a meaningful "fusion" policy which comprises all these behaviors. To this end, we propose four different policy fusion methods for combining pre-trained policies. We further demonstrate how these methods can be used in combination with Inverse Reinforcement Learning in order to create intelligent agents with specific behavioral styles as chosen by game designers, without having to define many and possibly poorly-designed reward functions. Experiments on two different environments indicate that entropy-weighted policy fusion significantly outperforms all others. We provide several practical examples and use-cases for how these methods are indeed useful for video game production and designers.