How an adaptive learning rate benefits neuro-fuzzy reinforcement learning systems

Abstract

To acquire adaptive behaviors of multiple agents in the unknown environment, several neuro-fuzzy reinforcement learning systems (NFRLSs) have been proposed Kuremoto et al. Meanwhile, to manage the balance between exploration and exploitation in fuzzy reinforcement learning (FRL), an adaptive learning rate (ALR), which adjusting learning rate by considering “fuzzy visit value” of the current state, was proposed by Derhami et al. recently. In this paper, we intend to show how the ALR accelerates some NFRLSs which are reinforcement learning systems with a self-organizing fuzzy neural network (SOFNN) and different learning methods including actor-critic learning (ACL), and Sarsa learning (SL). Simulation results of goal-exploration problems showed the powerful effect of the ALR comparing with the conventional empirical fixed learning rates.