Hybrid neurofuzzy networks are addressed in this paper. The network models have two basic structures, a fuzzy neural inference system and a neural network. The fuzzy system contains fuzzy neurons modeled through logic and and or operations processed via t-norms and s-norms, respectively. The neural network has nonlinear elements placed in series with the previous logical elements. The neural fuzzy inference system encodes a set of if-then rules and its recurrent multilayered structure performs

more »

... uzzy inference. Learning is based on an associative reinforcement learning to update second layer weights, and gradient search for output layer weights. The recurrent fuzzy neural network is particularly suitable to model nonlinear dynamic processes. Computational experiments with modeling of an unknown nonlinear process suggest that the hybrid fuzzy neural models are simpler, learning is faster, and that approximation errors are lower when compared with its counterparts.