组合SVM和决策树精确建立CCP点

SVM&DT precisely establing Critical Control Point

危害分析与关键控制点(HACCP)系统能有效保证食品质量安全,是食品行业安全生产流程的标准。在HACCP系统实施中,最关键的是精确定义系统的关键控制点(CCP):从HACCP系统的海量数据中选取有效数据来决策和定义CCP。本文组合支持向量机(SVM)和决策树(DT)来精确确定CCP。此方法基于SVM&DT在多类分类问题的精确简单特点和对多类划分的优越表现。应用此方法对杭州某水产养殖有限公司的历史数据进行验证和分析,发现CCP的自动挖掘精度大大提高,实现了虾苗验收、水产药物验收关键点控制问题。

Hazard Analysis and Critical Control Point（HACCP） system can effectively guarantee the quality of food safe and is the standard to the safety production process of food .In the implementation of HACCP system, the more is the precise definition of critical control point system（CCP） for HACCP.That is：Selecting useful data from the mass data of HACCP system to decide and definite ccps.In this paper, Support vector machine（SVM） and decision tree（DT） accurately determine CCPs.This method is based on SVM DT with a simple and precise classification of the characteristics and an excellent perfemance in Multi-Class division.Using this method to vivificate and analyze historical data of an aquaculture corporation.in Hangzhou, we find that the precision of CCPS ＇ s has been improved and realize the critical control point（CCP） in the process of acceptance check for shrimp of fisheries drug.