摘 要
近几年来，奖学金的评定工作成为高校学生工作的1个重要组成部分。在实施过程中，科学、完善的评价体系起着重要的作用。在现行的各种奖学金评定中主要采用的是传统的总分或平均分排名法，这两种方法都因专业、年级、考试试题的深度和难度及评卷教师不同，带来较多的不合理性。本文分析了目前学生奖学金评定方法中所存在的问题，对当前奖学金的评定方法进行改进，从极值标准化入手，确定模糊评定等级、隶属度函数、权重集，运用夹角余弦法建立模糊相似矩阵再改造为等价矩阵，最后进行聚类分析。克服了上述不合理性，取得了比较满意的结果。并与总分平均分法比较，阐明了前者比后者更有利于客观地评定出优秀学生，调动学生的学习积极性。从而使得奖学金的评定工作更趋于科学化、合理化。
关键词：奖学金评定；模糊聚类分析；夹角余弦法；模糊相似矩阵；
Abstract
In the resent years, the evaluation work of undergraduate scholarship is an important component of the students work. In the course of implementing, scientific, perfect appraisal systems play an important role. But in the evaluation of scholarships of different schools, the traditional evaluation is ranking students on the base of their total or average points brings much unreasonableness due to different majors grades curricula, the depth and difficulty of tests as well as marking teachers. This paper analyses the issues which exit in the evaluation work of undergraduate of issue. Starting from extremum standardization, the article defines ranks of fuzzy evaluation, subordinate function, weight sets, it establishes fuzzy similar matrices by applying angle cosine and changing them into equivalent matrices, then it makes cluster analysis. The produce overcomes the above mentioned unreasonableness and obtains satisfactory results. Moreover, it compare with total score and average score what the former is more advantage with judgment model students and bring activeness of students. In order to improve the method and make the judgment of scholarship more scientific and quantitative.
Keywords: scholarship judgment; fuzzy clustering analysis; angle cosine; fuzzy similar matrices;
