基于标准偏移量的学生成绩K-means聚类分析算法研究

Studying the k-means clustering analysis algorithm based on the standard offset of student achievement

针对利用距离差聚类分析算法分析学生学科成绩不够准确的问题,提出了利用标准差计算标准偏移量构建目标函数的K-means聚类分析算法,给出了初始聚类中心选取办法和算法的描述及处理流程;实验结果分析可得,利用标准偏移量构建的学生成绩K-means聚类分析算法,符合对学生成绩按学科类别聚类分析的特性要求;该算法能够较好的实现学科成绩高度关联属性的聚类分析结果。

For the clustering analysis algorithm using distance difference is not quite accurate for analyzing the scores of students, the K-means clustering analysis algorithm is put forward using standard offset to establish objective function, and the description of algorithm and the method of selecting the initial cluster centers are given as well. It is concluded that the K-means clustering analysis algorithm established by using standard offset conform to the unique requirement of clustering analysis for classification of student scores by subject.