摘要
设计基于集成学习的全云化健康大数据整合系统,实现健康大数据的高效率、高精度整合。管理员操作管理操作层实现系统控制、调控以及应用;大数据分析层通过ELM预测模型的参数单步预测方法获取融合多维参数信息的健康数据预测结果,并采用Bagging集成学习方法融合ELM预测模型获取高精度的强学习模型,实现差异多维全云化健康大数据的有效判读;通过全云化健康大数据整合层中的整合管理器以及整合运行引擎整合健康大数据,并通过调控层将整合后的健康大数据反馈到大数据资源层中,存储到该层中的临时数据库以及元数据库中,同时这些数据库中的数据为应用软件数据库提供数据调度服务。实验结果说明,该系统整合健康大数据的整合量和整合效率高,且具有较高的空间存储容量和并发数据处理性能。
A all-cloud health big data integration system based on integrated learning is designed to achieve the high efficiency and high precision integration of health big data.The administrator can perform operation management of the operation layer to realize the system control,regulation and application.The big data analysis layer is used to obtain the health data prediction results integrating multi-dimensional parameter information by means of the parameter single-step prediction method of the ELM prediction model,and the Bagging integrated learning method is used to integrate the ELM prediction model to obtain the high-precision strong learning model,so as to realize the effective judgment of the differentiated multi-dimensional all-cloud health big data.The integration manager and integrated running engine in the all-cloud health big data integration layer are used to integrate the health big data,and the integrated health big data is fed back to the big data layer by means of the regulating layer and stored in the temporary database and metadata database in the layer.The data in these databases is used to provide data scheduling service for the application software database.The experimental results show that the system has high integration capacity and efficiency in integrating health big data,as well as high spatial storage capacity and concurrent data processing performance.
作者
张喆
汤永利
ZHANG Zhe;TANG Yongli(Faculty of Science,Hong Kong Baptist University,Hong Kong 999077,China;Henan Polytechnic University,Jiaozuo 454000,China)
出处
《现代电子技术》
北大核心
2020年第22期173-176,180,共5页
Modern Electronics Technique
基金
国家密码管理局“十三五”国家密码发展基金密码理论项目(MMJJ20170122)。
关键词
健康大数据
整合系统
系统设计
集成学习
预测建模
数据存储
health big data
integration system
system design
integrated learning
prediction modeling
data storage
