【摘要】 首先,对深圳市2010-2012年的食品抽检数据进行整理,采用秩和比法将抽检结果量化和分档排序,并用支持向量机回归模型预测未来3年的食品质量总体呈上升趋势;然后,对食品产地做聚类分析,通过SPSS相关性分析,发现食品质量与抽检地无关;接着,将季节量化成温度和湿度,与食品质量做相关性分析,得出只与微生物显著相关的结论;最后,假设每年食品抽检的经费一定,确定每个食品领域每个季节在生产地的分配权重,合理分配抽检批次,使得检出的不合格数最大。更多还原

【Abstract】 The first aim of this paper is to predict the upward trend of food quality in the following three years using support vector machine(SVM)regression model.To do this,we sort sampling data of Shenzhen 2010-2010,and then rank and quantify the sampling result in terms of RSR methods.Secondly,the fact that food quality is independent of the sampling area,is found by clustering analysis for food producing area and the SPSS correlation analysis.When we quantify seasons as the temperature and humidity and then make a correlation analysis between food quality and seasons,we find the food quality is only closely related to microorganisms.In order to make the detection of the total number of unqualified is the largest,the method is that assuming the funds of food sampling in every year is fixed,we determine the allocation weight of each field and each season for every kind of food,and reasonablely distribute sampling batch.更多还原