《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (2): 457-462.DOI: 10.11772/j.issn.1001-9081.2021050871

• 数据科学与技术 • 上一篇    

可解释性有序聚类方法及其应用分析

高苏1, 鲍君忠2, 王昕1, 王利东1()   

  1. 1.大连海事大学 理学院,辽宁 大连 116026
    2.大连海事大学 航海学院,辽宁 大连 116026
  • 收稿日期:2021-05-27 修回日期:2021-07-25 接受日期:2021-07-26 发布日期:2022-02-21 出版日期:2022-02-10
  • 通讯作者: 王利东
  • 作者简介:高苏(1993—),女,山东济宁人,硕士研究生,主要研究方向:不确定决策;
    鲍君忠(1968—),男,辽宁营口人,教授,博士,主要研究方向:海事公约、风险评估、决策方法;
    王昕(1978—),女,辽宁铁岭人,副教授,博士,主要研究方向:粒计算、模糊信息处理;
    王利东(1979—),男,辽宁喀左人,教授,博士,主要研究方向:风险评价、粒计算、不确定决策。
  • 基金资助:
    国家自然科学基金资助项目(61803065)

Interpretable ordered clustering method and its application analysis

Su GAO1, Junzhong BAO2, Xin WANG1, Lidong WANG1()   

  1. 1.School of Science,Dalian Maritime University,Dalian Liaoning 116026,China
    2.Navigation College,Dalian Maritime University,Dalian Liaoning 116026,China
  • Received:2021-05-27 Revised:2021-07-25 Accepted:2021-07-26 Online:2022-02-21 Published:2022-02-10
  • Contact: Lidong WANG
  • About author:GAO Su, born in 1993, M. S. candidate. Her research interests include decision-making under uncertainty.
    BAO Junzhong, born in 1968, Ph. D., professor. His research interests include maritime convention, risk assessment, decision-making method.
    WANG Xin, born in 1978, Ph. D., associate professor. Her research interests include granular computing, fuzzy information processing.
    WANG Lidong, born in 1979, Ph. D., professor. His research interests include risk assessment, granular computing, decision-making under uncertainty.
  • Supported by:
    National Natural Science Foundation of China(61803065)

摘要:

针对管理决策领域中的等级分析问题,构建了面向语义可解释性的有序聚类方法。首先,在获得样本的优势度的基础上,结合模糊描述和K-modes聚类方法建立海员幸福感指数的有序聚类方法;然后,在公理模糊集框架下对有序聚类结果赋予相应的语义解释,以此形成一种从定量到定性的决策辅助方法;最后,以我国海员职业幸福感指数的9 175份有效调查问卷为研究样本,通过所构建的有序聚类方法得到海员职业幸福感指数的等级划分及其相应的语义描述,并分析了影响海员职业幸福感指数的内在原因。分析表明,所提方法不仅可以产生满足用户指定约束的有序聚类结果,而且聚类结果具有可解释性、可理解性,同时具有良好的辅助决策的价值。

关键词: 有序聚类, 海员幸福感, 可解释性, 语义描述, 模糊描述

Abstract:

For solving grade analysis problems in the field of management decisions, an ordered clustering method for semantic interpretability was proposed. Firstly, based on obtaining the dominance degrees of the samples, the fuzzy description and K-modes clustering method were combined to establish an ordered clustering method of Chinese seafarers’ vocational happiness indexes. Secondly, the corresponding semantic interpretation was assigned to the ordered clustering results under the framework of Axiomatic Fuzzy Set (AFS); thereby, forming a decision-making aid method for transforming the quantitative information into the qualitative description. Finally, taking the 9 175 valid questionnaires of Chinese seafarers’ vocational happiness indexes as the research samples, the constructed ordered clustering method was applied to obtain the grading results of the seafarers’ vocational happiness indexes as well as their semantic interpretation,and the factors influencing seafarers’ vocational happiness indexes were analyzed. The proposed method can produce ordered clustering results that satisfy user-specified constraints, and the results are interpretable, understandable, and have good value in assistant decision-making.

Key words: ordered clustering, seafarer vocational happiness, interpretability, semantic description, fuzzy description

中图分类号: