基于极限学习机的肉制品质量风险预测研究  

Prediction of meat product quality risk based on extreme learning machine

作　　者：汪颢懿 卞玉芳[2] 张瑞芳 王星云[3] WANG Hao-yi;BIAN Yu-fang;ZHANG Rui-fang;WANG Xing-yun(Network center,Beijing Technology and Business University,Beijing 100048,China;Nuclear and radiation safety center,Beijing 100082,China;School of Computer and Information Engineering,Beijing Technology and Business University,Beijing 100048,China;National Engineering Laboratory for Agri - product Quality Traceability,Beijing 100048,China)

机构地区：[1]北京工商大学网络中心,北京100048 [2]环境保护部核与辐射安全中心,北京100082 [3]北京工商大学计算机与信息工程学院,北京100048 [4]农产品质量安全追溯技术及应用国家工程实验室,北京100048

出　　处：《计算机仿真》2019年第10期413-418,共6页Computer Simulation

基　　金：国家重点研发计划项目(2016YFD0401205)

摘　　要：食品质量风险预警是民生保障中的重大问题,针对目前常用预警方法存在训练时间过长、精确度低等问题,提出了一种基于极限学习机(ELM)的重点食品安全风险预警模型.首先对国家食品安全抽检检测信息系统中肉制品的抽样检验数据进行预处理,从中提取特征数据并进行属性选择;其次,分别建立ELM和核极限学习机(KELM)下的重点食品安全风险预警模型,对分类特征数据进行分析,进而得出预警结果;最后,与采用back propagation (BP)神经网络,支持向量机(svm)所预警得出的结果进行对比,实验结果表明基于核极限学习机的食品安全风险预警模型在准确度与训练时间上都优于其他预警模型,对食品安全能够进行更有效预测,提升了食品安全质量监管的工作效率.Prediction of food product quality risk is a major problem in the residents‘ livelihood security.On account of the traditional methods lacking efficiency and accuracy,this paper proposed a prediction model of key food safety based on extreme learning machine(ELM).Firstly, this research preprocessed sampling data of meat products from national food safety sampling platform.From the original data, characteristic data were selected and classified into the sampling data. Secondly,the prediction models of key food safety were constructed by ELM and kernel extreme learning machine(KELM) separately.The characteristic data were experimented by ELM,KELM,Back Propagation(BP) Neural Network and Support Vector Machines(SVMs) Neural Network. The experimental results demonstrate the prediction model based on KELM is superior to other prediction models in accuracy and training time and it can predict and ensure the quality of food safety.

关 键 词：食品安全 极限学习机 核极限学习机 预测 预警模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]