基于进化规划的FCMBP模糊聚类改进方法

Evolutionary programming based FCMBP fuzzy clustering method

分析了基于摄动的模糊聚类方法(fuzzy clustering method based on perturbation,FCMBP),指出指数复杂度的遍历过程是目前PC计算环境下难以处理十阶以上较高阶数模糊相似矩阵的原因.把寻求具有最小"失真"的最优模糊等价矩阵看作优化问题来求解,提出了一种基于进化规划的FCMBP模糊聚类改进方法.与FCMBP相比,该方法通过引入基于进化规划的优化技术避免了遍历过程,使其能够对高阶模糊相似矩阵进行处理.得到的等价矩阵"失真"小于传递闭包法所得结果,从而获得更为精确可靠的聚类效果,将FCMBP模糊聚类方法推广到能够处理高阶模糊相似矩阵的情形,满足应用需要.

In current PC computational environment, the fuzzy clustering method based on perturbation （FCMBP） is failed when dealing with similarity matrices whose orders are higher than ten. The reason is that the traversal process in FCMBP is exponential complexity. This paper treated the process of finding fuzzy equivalent matrices with smallest error from an optimization point of view and proposed an improved FCMBP fuzzy clustering method based on evolutionary programming. Compared with FCMBP, the improved method can deal with high order matrices by introducing an evolutionary programming based optimization technique instead of the traversal process. A much more accurate solution could be obtained than that obtained by the transitive method. The improved method extends FCMBP to fit high order matrices, which meets the need of using FCMBP in real application problems.