摘要
传统的数据集成挖掘方法在集成与挖掘两个步骤之间存在较大误差,导致大数据出现乱码问题,数据显示不全。为解决上述问题,提出基于半监督深度学习法的大数据集成挖掘方法。利用有监督与无监督深度学习间的机器学习,组成半监督深度学习。利用支持向量数据组建立超球体。依据超球体结合标记样本,组建半监督深度学习数据检测模型,筛选样本特征词,利用半监督深度学习方法训练单分类SVDD模型,实现网络大数据集成挖掘。仿真结果证明,所提方法能够高精度、高效的对大数据完成集成挖掘,具有理想的应用性能。
The traditional data integration mining method has a large error between the two steps of integration and mining,which leads to the problem of garbled big data and incomplete data display.In order to solve the above problems,a big data integration mining method based on semi supervised deep learning is proposed.Firstly,semi-supervised deep learning was constructed.Secondly,according to the support vector data set,the hypersphere was founded.Then,semi-supervised deep learning data detection model was established to filter sample feature words based on the labeled samples combined with hypersphere.Finally,the semi supervised deep learning method was used to train the single class SVDD model,thus realizing the integrated mining of network big data.The simulation results show that the method has high precision and high efficiency of big data integration mining,and outstanding applicability.
作者
纪冲
刘岩
JI Chong;LIU Yan(College of Computer and Information,Inner Mongolia Agricultural University,Hohhot Inner Mongolia 010018,China)
出处
《计算机仿真》
北大核心
2021年第7期313-316,共4页
Computer Simulation
基金
国家自然科学基金(地区项目科研项目)(31960494)。
关键词
半监督学习
网络大数据
集成挖掘
检测模型
Semi-supervised learning
Network big data
Integrated mining
Detection model
