Timezone: »
Reinforcement learning (RL) in partially observable settings is challenging because the agent’s observations are not Markov. Recently proposed methods can learn variable-order Markov models of the underlying process but have steep memory requirements and are sensitive to aliasing between observation histories due to sensor noise. This paper proposes dynamic-depth context tree weighting (D2-CTW), a model-learning method that addresses these limitations. D2-CTW dynamically expands a suffix tree while ensuring that the size of the model, but not its depth, remains bounded. We show that D2-CTW approximately matches the performance of state-of-the-art alternatives at stochastic time-series prediction while using at least an order of magnitude less memory. We also apply D2-CTW to model-based RL, showing that, on tasks that require memory of past observations, D2-CTW can learn without prior knowledge of a good state representation, or even the length of history upon which such a representation should depend.
Author Information
Joao V Messias (Morpheus Labs)
Shimon Whiteson (Oxford University)
More from the Same Authors
-
2022 Poster: In Defense of the Unitary Scalarization for Deep Multi-Task Learning »
Vitaly Kurin · Alessandro De Palma · Ilya Kostrikov · Shimon Whiteson · Pawan K Mudigonda -
2022 Poster: Equivariant Networks for Zero-Shot Coordination »
Darius Muglich · Christian Schroeder de Witt · Elise van der Pol · Shimon Whiteson · Jakob Foerster -
2019 : Bayes-Adaptive Deep Reinforcement Learning via Meta-Learning - Invited Talk »
Shimon Whiteson