Representation-Induced Algorithmic Bias - An Empirical Assessment of Behavioural Equivalence over 14 Reinforcement Learning Algorithms Across 4 Isomorphic Gameform Representations.
Simon C. StantonJulian R. DermoudyRobert OllingtonPublished in: AI (2022)
Keyphrases
- reinforcement learning algorithms
- reinforcement learning
- state space
- model free
- markov decision processes
- reinforcement learning problems
- function approximation
- temporal difference
- learning algorithm
- reinforcement learning methods
- active learning
- eligibility traces
- neural network
- search space
- function approximators
- policy search