Posted May 24, 2019
Capacity Limited Reinforcement Learning in Minds and Machines
Summary

Exploit the cognitive science of “generalization” in human learning in order to develop improved AI algorithms.

Key Findings

Demonstrated Reinforcement Learning system that

  • Has improved generalization ability
  • Faster/more robust learning
Project Start Date
Project End Date

Publications

Lerch, R. A. & Sims, C. R. (2019a). "Rate-Distortion Theory and Computationally Rational Reinforcement Learning." Proceedings of Reinforcement Learning and Decision Making (RLDM) 2019, Montreal, Canada.

Malloy, T. J., Lerch, R. A., Fang, Z. & Sims, C. R. (2019). "Predicting Human Choice in a Multi-Dimensional N-Armed Bandit Task Using Actor-Critic Feature Reinforcement Learning." Proceedings of Reinforcement Learning and Decision Making (RLDM) 2019, Montreal, Canada.

Lerch, R. A. & Sims, C. R. (2018). "Exploration and policy generalization in capacity-limited reinforcement learning." Proceedings of the International Conference on Machine Learning (ICML) Workshop on Exploration in RL, Stockholm, Sweden.