智能系统学报
智能繫統學報
지능계통학보
CAAI TRANSACTIONS ON INTELLIGENT SYSTEMS
2015年
4期
627-635
,共9页
模糊C均值算法%聚类分析%遗传算法%动态分析%模糊聚类%初始值%避免早熟%全局最优%局部最优
模糊C均值算法%聚類分析%遺傳算法%動態分析%模糊聚類%初始值%避免早熟%全跼最優%跼部最優
모호C균치산법%취류분석%유전산법%동태분석%모호취류%초시치%피면조숙%전국최우%국부최우
fuzzy C-means clustering%cluster analysis%genetic algorithm%dynamic analysis%fuzzy clustering%initial values%premature contraction avoidance%global optimum%local optimum
针对传统的模糊C均值聚类( fuzzy Cm-eans clustering )算法容易陷入局部最优解,并且对初始值敏感的缺陷,提出一种基于改进的遗传算法的模糊聚类算法。该算法针对遗传算法的早熟问题提出一种改进的遗传算法,并将其应用于FCM算法,来寻找全局最优的聚类中心。实验表明,该算法与基于传统遗传算法的FCM算法相比,具有更强的寻优能力,更优的聚类效果。
針對傳統的模糊C均值聚類( fuzzy Cm-eans clustering )算法容易陷入跼部最優解,併且對初始值敏感的缺陷,提齣一種基于改進的遺傳算法的模糊聚類算法。該算法針對遺傳算法的早熟問題提齣一種改進的遺傳算法,併將其應用于FCM算法,來尋找全跼最優的聚類中心。實驗錶明,該算法與基于傳統遺傳算法的FCM算法相比,具有更彊的尋優能力,更優的聚類效果。
침대전통적모호C균치취류( fuzzy Cm-eans clustering )산법용역함입국부최우해,병차대초시치민감적결함,제출일충기우개진적유전산법적모호취류산법。해산법침대유전산법적조숙문제제출일충개진적유전산법,병장기응용우FCM산법,래심조전국최우적취류중심。실험표명,해산법여기우전통유전산법적FCM산법상비,구유경강적심우능력,경우적취류효과。
The traditional fuzzy C-means( FCM) clustering algorithm is prone to fall into the solution of local opti-mum and is sensitive to initial value.Aiming at these drawbacks, a fuzzy C-means based on the improved genetic algorithm is presented.The improved genetic algorithm is employed to optimise the FCM algorithm, finding the cluster center of the global optimum.Finally, the experimental results show that compared with the traditional FCM, the proposed algorithm has stronger optimisation ability and better clustering effect.