大数据精准挖据处理架构及预测模型研究  

New mining architecture and prediction model for big data

作　　者：杨斐[1] 艾晓燕[1] 张永恒[1] 张峰[1] 

机构地区：[1]榆林学院信息工程学院,陕西榆林719000

出　　处：《电子设计工程》2016年第12期29-32,共4页Electronic Design Engineering

基　　金：榆林学院科研项目(14YK38);榆林市科技计划项目(2014cxy-09)

摘　　要：为了提高大数据的精准挖据与预测能力,解决传统数据挖据技术无法适应大数据处理环境的问题,利用云计算和大数据处理技术,提出了大数据精准挖据处理架构及基于BP神经网络的预测模型。重点研究了大数据处理平台架构、大数据分析与表达技术、基于BP神经网络的大数据挖据及预测模型。应用结果表明,该方案结合云计算平台和大数据挖掘技术,能够高效的处理海量数据的处理和表达,对于销售数据具有一定的预测能力。In order to improve the accuracy of big data mining and forecasting ability, to solve the traditional data mining technology cannot adapt to big data processing environment problem, using of cloud services and big data processing technology, a new mining architecture and forecast model for big data model based on BP neural network is proposed. The structure of big data processing platform, big data analysis and expression technology and big data mining and prediction model based on BP neural network is designed. Application results show that the scheme combining cloud service platform and big data mining technology can effectively dealing with massive data processing and expression has a certain predictive ability for the sales data.

关 键 词：大数据 数据挖据 预测模型 BP神经网络 销售数据 

分 类 号：TN391[电子电信—物理电子学]