Count-Based Exploration in Feature Space for Reinforcement Learning

25 Jun 2017  ·  Jarryd Martin, Suraj Narayanan Sasikumar, Tom Everitt, Marcus Hutter ·

We introduce a new count-based optimistic exploration algorithm for Reinforcement Learning (RL) that is feasible in environments with high-dimensional state-action spaces. The success of RL algorithms in these domains depends crucially on generalisation from limited training experience. Function approximation techniques enable RL agents to generalise in order to estimate the value of unvisited states, but at present few methods enable generalisation regarding uncertainty. This has prevented the combination of scalable RL algorithms with efficient exploration strategies that drive the agent to reduce its uncertainty. We present a new method for computing a generalised state visit-count, which allows the agent to estimate the uncertainty associated with any state. Our \phi-pseudocount achieves generalisation by exploiting same feature representation of the state space that is used for value function approximation. States that have less frequently observed features are deemed more uncertain. The \phi-Exploration-Bonus algorithm rewards the agent for exploring in feature space rather than in the untransformed state space. The method is simpler and less computationally expensive than some previous proposals, and achieves near state-of-the-art results on high-dimensional RL benchmarks.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Atari Games Atari 2600 Freeway Sarsa-ε Score 29.9 # 34
Atari Games Atari 2600 Freeway Sarsa-φ-EB Score 0.0 # 57
Atari Games Atari 2600 Frostbite Sarsa-ε Score 1394.3 # 34
Atari Games Atari 2600 Frostbite Sarsa-φ-EB Score 2770.1 # 27
Atari Games Atari 2600 Montezuma's Revenge Sarsa-ε Score 399.5 # 22
Atari Games Atari 2600 Montezuma's Revenge Sarsa-φ-EB Score 2745.4 # 13
Atari Games Atari 2600 Q*Bert Sarsa-φ-EB Score 4111.8 # 46
Atari Games Atari 2600 Q*Bert Sarsa-ε Score 3895.3 # 47
Atari Games Atari 2600 Venture Sarsa-φ-EB Score 1169.2 # 18
Atari Games Atari 2600 Venture Sarsa-ε Score 0.0 # 49

Methods


No methods listed for this paper. Add relevant methods here