摘要
为了提高数据库中存储的数据量,提出了基于双链量子遗传优化的分类规则挖掘算法,并建立流数据挖掘平台,采用垂直并行化的方式处理源数据,提取符合用户需求的有用数据。从量子位实数编码、解空间变换等过程描述分类规则集形成并实现分类过程,利用目标函数的梯度信息调整进化步长幅值,以避免陷入局部最优解。流数据挖掘平台的数据采集系统中对ETL模型进行优化,提供的消息分发模型的限制下完成下行通信消息调度分发策略的设计,使数据采集系统具有双向通信的能力。通过试验,本研究流数据挖掘平台的数据处理时间最短,对实验数据集中Iri数据集的分类挖掘精度高达95.17%。
In order to increase the amount of data stored in the database,a classification rule mining algorithm based on double-strand quantum genetic optimization is proposed,and a streaming data mining platform is established.The source data is processed in a vertical parallel manner to extract useful data that meets the needs of users.From the process of real qubit encoding and solution space transformation,the formation of the classification rule set is described and the classification process is realized,and the gradient information of the objective function is used to adjust the evolution step amplitude to avoid falling into the local optimal solution.The ETL model is optimized in the data collection system of the streaming data mining platform,and the downlink communication message scheduling and distribution strategy design is completed under the limitation of the message distribution model provided,so that the data collection system has the ability of two-way communication.Through experiments,the data processing time of the streaming data mining platform in this study is the shortest,and the classification and mining accuracy of the Iri data set in the experimental data set is as high as 95.17%.
作者
杨恒
李心愉
Yang Heng;Li Xinyu(Haihe River,Huaihe River and Xiaoqinghe River Basin Water Conservancy Management and Service Center of Shandong Province Jinan 250100,China)
出处
《现代科学仪器》
2022年第5期206-212,共7页
Modern Scientific Instruments
关键词
双链量子遗传算法、流数据挖掘平台
实数编码
解空间变换
消息调度分发
double-chain quantum genetic algorithm,stream data mining platform
Real coding
Solution space transformation
Message scheduling distribution
