This article proposes a method of reinforcement learning (RL) of an interpretable fuzzy system (IFS) based on a neural fuzzy actor-critic (RLIFS-NFAC) framework.