期刊
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
卷 125, 期 -, 页码 785-791出版社
ELSEVIER
DOI: 10.1016/j.future.2021.07.025
关键词
Poset; Hasse graph; Hierarchical hierarchy; Cluster analysis
资金
- Key Projects of National Key R&D Plan in the 13th Five-Year Plan of China [2019YFD0901605]
This paper proposes a hierarchical clustering method based on posets to address the hierarchical classification issue in traditional evaluation. By converting traditional evaluation into partial evaluation and incorporating evaluators' preferences, the method successfully stratifies the clustering process, offering a coarse particle approach for evaluation grading.
In order to solve the hierarchical classification problem in traditional evaluation, a hierarchical clustering method is constructed by applying posets to the case of unknown grading standards. Firstly, the traditional evaluation is converted into partial evaluation by partial order relation, and the classification is implemented according to partially-ordered Hasse graph. Then, the partially-ordered Hasse graph is embedded with the preference of evaluators. Next, the information within the layer expresses the result of clustering, and the information between the layers reflects the difference of grades. After that, the structural property of the partially-ordered Hasse graph is proved. Although the traditional evaluation can be upgraded to partial order evaluation theoretically, the transformation mode of the two is still in the process of improvement. This paper solves the problem that traditional clustering cannot be stratified, and provides a coarse particle method that can be used to implement grading for evaluation. (C) 2021 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据