基于Weka和协同机器学习技术的数据挖掘方法研究  

Research on Data Mining Method Based on Weka and Cooperative Machine Learning Technology

作　　者：谭成兵[1] 周湘贞 朱云飞 TAN Chengbing;ZHOU Xiangzhen;ZHU Yunfei(Department of Intelligent Engineering, Bozhou Vocational and Technical College, Bozhou 236813, China;Institute of Financial Strategy, Chinese Academy of Social Sciences, Beijing 100028, China;Tsinghua University Press, Beijing 100084, China)

机构地区：[1]亳州职业技术学院智能工程系,安徽亳州236813 [2]中国社会科学院财经战略研究院,北京100028 [3]清华大学出版社,北京100084

出　　处：《长春大学学报》2020年第12期5-9,共5页Journal of Changchun University

基　　金：国家重点研发计划资助(2017YFF0106407);2017国家自然科学基金青年基金项目(61702026);河南省2018年度科技攻关项目(182102110277);安徽省高校自然科学研究重点资助项目(KJ2018A0881);安徽省高校优秀青年人才支持计划重点项目(gxyqZD2016557)。

摘　　要：为了提高数据挖掘准确度,提出了一种基于Weka和协同机器学习技术的数据挖掘方法。将Weka平台中具有相同训练结果的算法分至同组,且给每个算法得到的结果赋予权重因子。然后,采用蛙跳算法对权重因子进行训练,将训练得到的权重因子按照不同的组进行求和。最后将最大的一组所对应结果作为样本的训练输出,以实现不同机器学习算法协同。结果表明,相比于单个机器学习算法的数据挖掘,采用协同机器学习的数据挖掘准确度更高,稳定性更强。In order to improve the accuracy of data mining,a data mining method based on Weka and collaborative machine learning technology is proposed.The algorithms with the same training results in Weka platform are divided into the same group,and the results obtained by each algorithm are given weight factors.Then the weight factors are trained by frog leaping algorithm,and the weight factors obtained by the training are summed up according to different groups.Finally,the largest group of corresponding results is taken as the training output of samples to realize the cooperation of different machine learning algorithms.The results show that,compared with data mining based on single machine learning algorithm,the data mining based on collaborative machine learning has higher accuracy and stronger stability.

关 键 词：WEKA平台 协同机器学习 数据挖掘 权重因子 蛙跳算法 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]