Online fuzzy Q-learning with extended rule and interpolation technique

Q-learning is a kind of reinforcement learning where the agent solves the given task based on rewards received from the environment. Most research done in the field of reinforcement learning has focused on discrete domains. But the environment with which the agent must interact is continuous. Thus a method that is able to make Q-learning applicable to the continuous problem domain is needed. In the paper, the basic fuzzy rule is extended so that it can incorporate Q-learning. The interpolation technique which is widely used in memory-based learning is adopted to represent the appropriate Q-value for current state and action pair. The resulting structure based on the fuzzy inference system has the capability of solving the continuous state and action problem in Q-learning and generating fuzzy rules via interacting with the environment without a priori knowledge about the environment. The effectiveness of the proposed structure is shown through simulation on cart-pole system.