Fuzzy adaptive Q-learning method with dynamic learning parameters
Y. Maeda · 2002
An active search in the reinforcement learning disturbs the learning process when learning proceeds and converges to a partial search area. Therefore, it is important to balance between searching behaviors of the unknown knowledge and using the behavior of the obtained knowledge. In this research, we propose an adaptive Q-learning method for tuning the learning parameters of reinforcement learning by fuzzy rules. We also report the results of artificial ants simulation using this method.