摘要
将数据挖掘中凝聚的层次聚类思想引入矩形排样问题求解中。首先计算矩形件两两之间的结合度,将结合度最高的两个矩形件合并形成一个新的矩形件,并重复执行这一步骤,直到所有矩形件合并完成。据此编写算法程序,利用标准数据集进行仿真实验。
This paper introduces the idea of hierarchical clustering in data mining field to solve the rectangular pack- ing problem. First it calculates the combination degree between each two rectangular pieces, combines the two rectan- gles into a new one, which have the highest combination degree, and repeat this step until all the rectangles merge into one. The paper designs the procedure of algorithm and uses standard data set to get simulation experiment.
出处
《重庆科技学院学报(自然科学版)》
CAS
2014年第2期127-129,142,共4页
Journal of Chongqing University of Science and Technology:Natural Sciences Edition
基金
安徽省教育厅自然科学基金(KJ2013B232)
合肥学院国家级大学生创新创业训练计划项目(20121105902
201211059024
201311059105
201311059106)
安徽省自然科学基金(1308085QF115
1308085MF84)
关键词
矩形排样
层次聚类
结合度
特征矩阵
rectangular packing problem
hierarchical clustering
combination degree
characteristic matrix