基于均值平移和区间变换组合法的公平化处理算法  

A Fairness Processing Algorithm Based on the Combination of Average Value Regulation and Range Transformation

作　　者：杨平[1] 马进明[1] 李平[1] 胡隽[1] 

机构地区：[1]上海电力学院电力与自动化工程学院,上海200090

出　　处：《上海电力学院学报》2008年第1期84-86,97,共4页Journal of Shanghai University of Electric Power

摘　　要：讨论了多名教师对多组学生作业评分成绩的公平化数据处理问题.提出了3种实施性较强的算法:均值平移法、区间变换法和组合法.研究表明:用均值平移法可使组均值等于整体均值,但组分散区间与整体分散区间的差较大,还存在超限处理会产生新的不公平的问题;用区间变换法可使组分散区间等于整体分散区间,但组均值与整体均值的差较大;用组合法既使组分散区间等于整体分散区间,而且又使组均值向整体均值靠拢,他们之间的差比区间变换法的要小一半.组合法综合了均值平移法和区间变换法的优点,值得推荐.The fairness processing problem of group student scores is discussed here. Three exercisable algorithms, average value regulation, range transformation and the combination methods, are presented. The study shows that with average value regulation, group average value equalizes overall average value, but group range differs from overall range, and probably new unfair problems occur at over upper limit, with range transformation, group range equalizes overall range, but group average value differs from overall average value; with those combination methods, group range equalizes overall range and group average value follows overall average value, with only half as many errors as in the case of range transformation method. Therefore the combination method takes the advantage over average value regulation and range transformation methods.

关 键 词：数据处理 统计分析 公平 均值 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]