欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了解决模糊C均值聚类算法（FCM）中聚类类数初始值是由先验知识人为确定并且目标函数忽略了样本属性数据之间的不均衡性问题，提出了一种基于模拟退火的样本加权FCM算法（SASWFCM），利用模拟退火算法可以寻求全局最优解的特点，计算出聚类数初始值，并对聚类中心和目标函数进行加权处理。通过实验分析，该算法与原FCM算法相比较而言，无需人为确定聚类初始值并且在分类准确数和准确率上有所提高，体现了算法的优越性，验证了改进后算法的实际价值。
英文摘要：
When using fuzzy C-means clustering algorithm, the initial value of the number of clustering classes needed to be determined by prior knowledge of people and the objective function ignored the disequilibrium problems in sample attribute data. This paper put forward a kind of sample weighted FCM algorithm based on simulated annealing algorithm. It calculated the number of clustering initial value using the simulated annealing algorithm which has an excellent feature of solving globally optimal solution problems and pressed weighted processing on the clustering center and the objective function. The experimental analysis shows that this algorithm needs not determine the number of clustering initial value by prior knowledge of people compared to other FCM algorithms and it is superior to the traditional FCM algorithm at classification accuracy and classification accuracy rate.
同期刊论文项目
同项目期刊论文
期刊信息
位置：