An important property for lifelong-learning agents is the ability to combine existing skills to solve unseen tasks. In general, however, it is unclear how to compose skills in a principled way. We provide a "recipe" for optimal value function composition in entropy-regularised reinforcement learning (RL) and then extend this to the standard RL setting. Composition is demonstrated in a video game environment, where an agent with an existing library of policies is able to solve new tasks without the need for further learning.
@article{arxiv.1807.04439,
title = {Will it Blend? Composing Value Functions in Reinforcement Learning},
author = {Benjamin van Niekerk and Steven James and Adam Earle and Benjamin Rosman},
journal= {arXiv preprint arXiv:1807.04439},
year = {2018}
}
Comments
The 2nd Lifelong Learning: A Reinforcement Learning Approach (LLARLA) Workshop, Stockholm, Sweden, FAIM 2018