We consider a general approach to knowledge transfer in which an agent learning with a neural network adapts how it reuses existing networks as it learns in a new domain. Networks trained for a new domain are able to improve performance by selectively routing activation through previously learned neural structure, regardless of how or for what it was learned. We present a neuroevolution implementation of the approach with application to reinforcement learning domains. This approach is more general than previous approaches to transfer for reinforcement learning. It is domain-agnostic and requires no prior assumptions about the nature of task relatedness or mappings. We analyze the method's performance and applicability in high-dimensional Atari 2600 general video game playing.

PDF

In IJCAI'15 Workshop on General Intelligence in Game-Playing Agents, pp. 7--14 2015.

@inproceedings{braylan:ijcai15ws, title={On the Cross-Domain Reusability of Neural Modules for General Video Game Playing}, author={Alexander Braylan and Mark Hollenbeck and Elliot Meyerson and Risto Miikkulainen}, booktitle={IJCAI'15 Workshop on General Intelligence in Game-Playing Agents}, pages={7--14}, url="http://www.cs.utexas.edu/users/ai-lab?braylan:ijcai15ws", year={2015} }