基于模糊软集合的缺失数据新的预测方法  

A New Prediction Method of Missing Data Based on Fuzzy Soft Set

作　　者：孔芝[1] 赵杰[1] KONG Zhi;ZHAO Jie(College of In formation Science and Engineering,Northeastern University,Shenyang 110819»China)

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110819

出　　处：《模糊系统与数学》2021年第5期152-163,共12页Fuzzy Systems and Mathematics

基　　金：河北省自然科学基金资助项目(F2017501041)

摘　　要：数据缺失现象是无法避免的客观存在的问题,数据填充是一种很好的处理该问题的方法。数据的类型多种多样,很多数据含有不确定性,模糊软集合理论是有效的处理不确定问题的数学工具。在利用该理论对缺失数据进行填充的研究中,现有方法准确度不高,且某些情况下填充值超限。为解决此问题,本文提出一种新的模糊软集合缺失数据的预测方法。利用对象与参数间的关系,根据平均距离,计算参数权重和对象权重,由最相关的参数和对象数据填充缺失数据。为验证方法的有效性,选取UCI数据库进行实验,结果表明所提出的方法比现有方法更加准确和有效。Data missing is an inevitable and actually existing problem. Data filling is a good method to deal with this problem. There are various types of data, and some data are uncertain and missing. Fuzzy soft set theory is an effective mathematical tool to deal with uncertain problems. In filling missing data by fuzzy soft set theory, the accuracy of existing methods is not high, and the filling values exceed the range in some cases.In order to solve these problems, a new prediction method is proposed based on fuzzy soft set. Using the relationship between objects and parameters, the parameter weight and object weight are calculated according to the average distance, and the missing data is filled with the most relevant parameters and object data. In order to verify the effectiveness of the new method, some data sets from UCI database are selected for experiments. The results show that the proposed method is more accurate and effective than the existing methods.

关 键 词：模糊软集合 缺失数据 绝对平均距离 熵测度 平均准确率 

分 类 号：O159[理学—数学]