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：
与通常的聚类方法相比，客观聚类分析方法能自动、客观地确定聚类个数并找到最优聚类方案．通过算法步骤的剖析和算例分析，对客观聚类分析方法的核心构件，即一致性准则的不足进行了评价．利用偶极子给出了新的一致性准则，提出了数据分组处理（GMDH）聚类分析方法．从理论分析和实证比较两方面论证了新的一致性准则的优越性，指出了GMDH聚类分析方法是客观聚类分析方法的发展．
英文摘要：
Compared with the usual clustering method, the objective cluster analysis can automatically and objectively determine the number of clusters and find out the optimal clustering scheme. The shortcoming of the consistency criterion of the objective ctuster analysis, that is its core component, is appraised by analyzing its algorithm steps and examples. Using the dipole concept, a new consistency criterion is given and a group method of data handling （GMDH） cluster analysis approach is presented. The advantage of the new consistency criterion is proved through theoretical analysis and demonstration comparisons. The results show that the GMDH cluster analysis is the further development of the objective cluster analysis.
同期刊论文项目
同项目期刊论文
期刊信息
位置：