摘要
对猪肉、牛肉、羊肉及虾等几种生鲜农产品进行了减压贮藏实验,通过检测各种样品不同保藏时间的挥发性盐基氮含量(TVB-N)、细菌总数、pH值及感官评分数据,以期实现对其新鲜度的准确分类。实验结果表明,任何单一理化或感官指标都难以获得理想的分类正确率。在此基础上,运用支持向量机(support vector machine,SVM)方法对以上数据进行合理的综合训练,并对参数进行优化,从而得到SVM神经网络模型,利用此模型进行肉品的新鲜度分类预测,可大大提高分类正确率。
Several fresh agricultural products, including pork, beef, mutton and shrimp samples, were stored in decompression storeroom, and the TVB-N content, total bacterial count, pH value and sensory scores of these samples in different time were determined to achieve the correct classification of freshness. The experiments showed that it was difficult to obtain the ideal classification accuracy by any single physicochemical or sensory properties.Therefore, SVM was taken into consideration to train the experimental data and the parameters would be optimized by rough and precise selection.And the obtained SVM model could be used to Dredict the meat freshness with high classification accuracy.
出处
《食品工业科技》
CAS
CSCD
北大核心
2011年第4期112-116,共5页
Science and Technology of Food Industry
基金
上海市晨光计划项目(2008CG055)
上海市晨光计划项目(09CG50)
关键词
支持向量机
肉品新鲜度
分类
support vector machine
meat freshness
classification
