Improved self-organizing CPN-based fuzzy system

An improved self-organizing CPN-based fuzzy system is proposed in this paper. Associated with the neuro-fuzzy system, there is a two-phase hybrid learning algorithm, which utilizes a CPN-based nearest-neighborhood clustering scheme for both structure learning and initial parameters setting, and a gradient descent method with variable learning rate for parameters fine-tuning. By combining the above two methods, the learning speed is much faster than that of the original back-propagation algorithms. The comparative results on the examples suggested that the method has the merits of simple structure, fast learning speed and good modeling accuracy.