An Improved Method on Nearest-neighbor Clustering Learning in Self-adaptive Fuzzy Identification

A kind of method on self-adaptive fuzzy identification based on fuzzy nearest-neighbor clustering was analysed and two shortcomings were explored. This paper proposes the corresponding improvements. Dividing the data space into high-density area and low-density area, adjusting the centers of each class, both of the improvements help the fuzzy model approximate the nonlinear system better. Simulation results show the advantage of the improvements