Abstract
In this paper, we introduce a general category of multi-fuzzy-neural networks (FNNs), analyze their underlying architecture and propose a comprehensive identification framework. The proposed multi-FNNs dwells on a concept of linear fuzzy inference-based FNNs. The design of the model uses a standard HCM (Hard C-Means) clustering algorithm and carries out an evolutionary fuzzy granulation of experimental data. The performance of the model is quantified through a series of experimental studies involving synthetic and real-world data.