报告人：鲜思东
报告地点：数学与统计学院415室
报告时间：2019年09月22日星期日10:30-11:30
邀请人：郑术蓉
报告摘要：
For observations to be classified, when scoring rules are imprecise or the cost of their computation is too high, the clustering method under linguistic information is necessary. Considering the accuracy of intuitionistic fuzzy linguistic variable in expressing experts’ opinions, a clustering algorithm is presented in this paper. Firstly, the concept of triangular intuitionistic fuzzy linguistic variables (TIFLVs) is introduced, and a new formula is developed for calculating correlation coefficient of TIFLVs. Then, the correlation coefficient plays a central role in our modified λ-cutting algorithm for clustering, which is utilized to construct an equivalence correlation matrix. In addition, a silhouette cluster validity index of TIFLVs is proposed to revise the results of clustering. Finally, the experimental results demonstrate the application and practicability of the clustering algorithm.
主讲人简介：
鲜思东 ，男，教授，硕士生导师，复杂系统智能分析与决策重庆市高校重点实验副主任,重庆邮电大学数据分析与智能决策创新团队负责人,中国商业统计学会理事，CSIAM大数据与人工智能专业委员会委员，中国现场统计学会大数据统计分会常务理事，中国商业统计学会数据科学与商业智能分会常务理事，重庆市工业与应用数学学会理事、副秘书长，重庆现代智慧教育研究院副院长（兼），国家教学成果奖评审专家，重庆市质量大数据分析应用决策咨询专家，重庆市科学道德和学风建设宣讲专家 。 研究方向为：统计分析与建模、语言信息建模、大数据与智能决策、模糊时间序列分析、供应链管理、不确定信息优化与决策等。现主持“质量大数据分析应用平台建设总体设计”重大委托项目1项等，“大数据视阈下重庆精准脱贫机制改革研究”重庆市社科规划项目1项，曾主持参与国家及省部级以上项目7项。在Journal of Optimization Theory and Applications 、International Journal of Intelligent Systems 、Soft Computing等国内外重要学术期刊发表相关论文70多篇，其中SCI论文21篇，SSCI论文5篇，EI论文19篇，CSSCI论文4篇等。获省部级以上教学或科研成果奖励5项。主编概率论与数理统计等教材3部。现担任国际杂志Transactions on Systems, Man, and Cybernetics: Systems(SCI)，International Journal of Systems Science(SCI,EI), Knowledge-Based Systems(SCI)，International Journal of fuzzy systems(SCI), Soft Computing (SCI)等多个权威期刊的审稿人。
