automatic fuzzy clustering

In this paper, the automatic fuzzy clustering is presented in multiobjective optimization framework. Two objectives based on compactness and overlap-separation measures are used as the objective functions which are optimized simultaneously. The proposed optimization problem is a nonlinear and non-convex multi-objective optimization and accordingly, a grey wolf based optimization algorithm is proposed for solving it. For accelerating the optimization process and preventing local optimum trapping, new heuristic approaches are included to the original algorithm. Solving the multi-objective optimization problem using the proposed algorithm, results in several Pareto optimal solutions showing compromise between objective functions. The final solution, among the generated plans, is selected using DB cluster validity index. The proposed method is applied on synthetic and real-life data sets. As shown in the experiments, the approach provides promising solutions in well-separated, hyperspherical and overlapping clusters. This is demonstrated by the comparison with existing single-objective and multiobjective clustering techniques.