朴素贝叶斯分类模型在中红外光谱测原料奶掺假中的应用研究  

Application research on naive bayesian classification model in detecting adulterated composition in milk

作　　者：张微[1,2] 

机构地区：[1]河北农业大学,保定071001 [2]河北省产品质量监督检验院,石家庄050051

出　　处：《食品工程》2011年第1期49-53,共5页Food Engineering

摘　　要：现有常规乳与乳制品掺假检测方法主要是利用鲜乳以及所掺入物质的物理化学性质进行测定,每种方法只能检测一种掺假成分,所以比较费时、烦琐。中红外光谱具有分析快速、操作方便、成本低、对样品不造成损伤、无需前处理、再现性好、不污染环境等优点。通过朴素贝叶斯分类模型建立生鲜牛奶和掺假原料奶的判别模型,判断未知样是否属于掺假原料奶。Measuring the chemical and physical properties of fresh milk and the mixing material is mainly used to detect the milk and adulterated milk.Each method can detect one adulterated composition only,so it is time-con-suming and cumbersome.Rapid MIR analysis has advantages such as convenient operation,low cost,no damage to sample pretreatment,reproducibility good,no pollution,etc.It is primarily through Naive Bayesian Classification Model to establish discriminant model of raw milk and adulterated milk to judge unknown samples.

关 键 词：鲜乳 掺假 中红外光谱 朴素贝叶斯分类模型 

分 类 号：TS252.7[轻工技术与工程—农产品加工及贮藏工程]