DMGA-FCM:衍生多种群遗传进化的FCM自适应聚类算法  

DMGA-FCM:Adaptive FCM Based on Derived Multi-population Genetic Algorithm

作　　者：冯志豪 曹金鑫 黄嘉爽 鞠恒荣 程纯[1] 丁卫平[1] FENG Zhi-hao;CAO Jin-xin;HUANG Jia-shuang;JU Heng-rong;CHENG Chun;DING Wei-ping(School of Information Science and Technology,Nantong University,Nantong 226019,China)

机构地区：[1]南通大学信息科学技术学院,江苏南通226019

出　　处：《小型微型计算机系统》2023年第10期2196-2203,共8页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61976120,62006128,62102199)资助;江苏省自然科学基金项目(BK20191445)资助;江苏省双创博士计划江苏省高等学校自然科学研究重大项目(21KJA510004)资助;江苏省高等学校自然科学研究面上项目(21KJB520018,20KJB520009)资助;教育部人文社会科学研究青年基金项目(21YJCZH013)资助;南通市科技局基础科学研究项目(JC2020141,JC2021122)资助.

摘　　要：模糊C均值聚类(Fuzzy C-means Clustering,FCM)算法是分析医学数据的重要方法之一,FCM的聚类效果容易受初始聚类中心的影响;诸多研究人员往往采用多种群遗传算法(Multiple Population Genetic Algorithm,MPGA)解决上述问题,但MPGA的全局搜索能力不足并缺少自适应性、易过早收敛、初始聚类中心不佳.为此,本文提出一种DMGA-FCM:衍生多种群遗传进化(DMGA)的FCM自适应聚类算法.在DMGA-FCM中,本文首次提出的衍生算子,对初始化种群进行衍生操作,提升算法寻优能力,处理种群间寻优能力不足;利用模糊控制动态调节遗传概率,以提升算法自适应性,进而增强DMGA算法全局寻优能力,避免过早收敛;用DMGA优化FCM算法的初始聚类中心,以提升算法聚类效果.在仿真实验中,本文将该算法与其他相关FCM算法进行对比,可得到更优的医疗数据聚类效果和图像聚类分割效果.Fuzzy C-means Clustering algorithm(FCM)is one of the important methods to analyze medical data,and the clustering effect of FCM is easily affected by the initial clustering center.Many researchers often use Multiple Population Genetic Algorithm(MPGA)to solve this problem.Still,MPGA has insufficient global search capabilities and lacks self-adaptability,so it is prone to premature convergence,and the initial cluster center is not good.For this reason,we proposed an DMGA-FCM:adaptive FCM clustering algorithm based on Derived Multi-population Genetic Algorithm.The first proposed derivation operator is used for DMGA-FCM,which takes into account the problem of insufficient optimization ability among populations.The operator performs derivative operations on the initial population to improve the optimization ability of the algorithm.Meanwhile,the genetic probability is dynamically adjusted using fuzzy control to improve the algorithm adaptability,thereby enhancing the global optimization capability of the DMGA algorithm and avoiding premature convergence problems.Then it is used to optimize the FCM algorithm to improve the clustering effect of the algorithm.The results of simulation experiments show that this algorithm can get better medical data clustering effect and image clustering segmentation effect compared with other related FCM algorithms.

关 键 词：模糊C均值聚类 多种群遗传算法 衍生算子 模糊控制 聚类中心 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]