Pickles Sampling Results Prediction Model And Its Application Of Yuhang District

BackgroundPickles refers to the fresh vegetables as the main raw material, produced using different techniques from a variety of salted vegetable products in general. Pickles has a long history, not only delicious, but contains many nutrients of its own, has always been loved by the people.In Yuhang District, it is the people standing breakfast thing. How the quality of pickles in Yuhang District, is directly related to people's daily life and human health. Therefore, it is necessary to study the sampling results of Yuhang District, pickles prediction model, and apply the model for the 2011 sampling results to predict pickle proposed sampling warning.Materials and MethodsIn this study, use of Yuhang District Health Authority 2001-2010 pickles registration data sampling results, the following studies:①determine the effects of sampling results in Yuhang District, pickles factors. Sampling units and production units from two aspects, the integration of primary health supervision departments sampling of the actual situation of pickles, Yuhang District,2001-2010 sampling of the original data, determine the sampling that may affect the results of Yuhang District, pickles factors; ②Construction of Yuhang District, pickles sampling results forecasting model. Using binary logistic regression analysis as research methods, a variety of factors and sampling pickles, Yuhang District, whether the association between the results of statistical significance, and build the results of Yuhang District, pickles sampling prediction model;③the prediction model application. Application of predictive model has been constructed, pickles, Yuhang District, conducted in 2011 predicted the results of sampling and analysis of projections, made pickles of sampling warning.Results1 2001-2010 pickles sampling results in Yuhang District Overview2001-2010, a total of 236 samples of sampling pickles, of which 78 failed, accounting for 33.05%; pass 158, accounting for 66.95%. Since 2001, Yuhang District, pickles pass rate gradually increased sampling to maintain the oscillation after 2004 than in the middle and upper levels. Sampling a minimum pass rate in 2001 was 31.82%. Sampling the highest pass rate in 2005, was 92.50%; sampling pass rate in 2005 reached its peak after the oscillations began to appear, there are two bottom of a sub-peak.2010 91.67% pass rate sampling, but the sampling results trend is unclear. Food Additives project as the main failure factors, proportion is 93.18%. Pass rate in different years of sampling the difference was statistically significant.2 According to the passing rate of distribution factorsNot carry out the work of removing sampling in February and November, and the number is only a sampling of June, Yuhang District, pickles work sampling sampling at 1 month pass rate the highest,85.71%; August of sampling pass rate lowest,22.72%; July, August, September pass rates for three consecutive months of sampling at a lower level of maintenance; pass between the different sampling rate of the month the difference was statistically significant. Food processing units and the sampling pickles higher pass rates,93.33% and 91.67%; food stalls of sampling minimum pass rate for 49.38%; different sampling units by type of sampling between the statistically significant have difference in pass rate. Liangzhu groups of sampling compliance rates for 75.61%; Linping pass rate sampling, Vice City, the lowest,63.95%; different sampling units were the pass rate differences between regions were not significant. The maximum sampling rate of large enterprise qualifying for 100%; small businesses the lowest pass rate sampling, for 66.15%; by decreasing the scale of production units, and its products is also decreasing in turn pass rate sampling; difference between the scale of production units was no statistical difference in the pass rate significance. Production units located in Yuhang District of sampling a pass rate of 76.47%, is located in Yuhang District of the sampling of non-compliance rate was 66.50%; production units are located in Yuhang District of the sampling was not statistically significant differences in passing rates. Production units located in Hangzhou, a pass rate of 55.56% of the sampling, in Hangzhou, the sampling of non-qualified rate 71.01%. Production units are located in Hangzhou, the sampling was statistically significant difference in pass rate. Production units located in Zhejiang Province, sampling a pass rate of 64.25%, in the pass rate of non-sampling Zhejiang Province 86.67%; production units are located in Zhejiang Province sampling statistically significant difference in pass rate.3 Binary logistic regression analysisAccording to a single factor non-conditional logistic regression analysis, sampling year, sampling units by type, scale of production units, production units are located in Yuhang District, Hangzhou, production units are located in Yuhang District of the five variables between the sampling results pickles association was statistically significant; and sampling month, sampling unit area is, the production units are located in Zhejiang Province, Yuhang District, the three variables and the relationship between the results of sampling pickles, not statistically significant. Yuhang District, pickles sampling results for the prediction model According to prediction model's -2 log likelihood value, model prediction accuracy, ROC curve of the situation, we could predict the general effect of model fitting.4 Application of the modelYuhang District in 2011 as the model prediction equation pickle sampling combined total of 9 variables. P2011 the whole series was a negative skewed distribution. Set if the P2011＜0.86, then the possibility of a large pickle sampling failed; as 0.86≤P2011≤0.99, then the possibility of passing the general sampling pickles; if P2011＞0.99, then the sampling of qualified pickles great possibilities. Sampling work in 2011, Yuhang District, the region should increase the focus on small-scale production of Xiaoshan District, product sampling efforts pickles. River Town, Xiaoshan District, in particular small businesses, they should take precautions to enhance sampling in order to ensure consumer safety aspects of food pickles. Within the region in Yuhang District, should avoid sampling Chongqing Fuling mustard (Group) Co., Ltd. product, unless really necessary.DiscussionHigher than the sensitivity of forecast model specificity, positive predictive value is also higher than the negative predictive value, so the passing rate of the model predicted a higher return is big pickle products. Sort=0.86, the prediction model's overall accuracy rate decline, although at the expense of correct prediction rate of the overall price, but the sampling of non-conforming products, forecast accuracy has been improved and increased revenue in this regard. Therefore, this sampling work in the pickle it is of practical significance.Regulatory functions of the health administration department of classes reduced from the original 6 to 1 class. Regulatory functions of the health administration department will change the sampling results, Yuhang District, pickles impact factors, which affect the prediction model prediction. Scale production units are too strict criteria for the classification prediction model will affect the unpredictable.The scale of production units by the lack of a development perspective, may cause partial Prediction Model Poor and the model prediction error.Conclusions1 Sampling year, sampling unit is the type of scale of production units, production units of the regional one (it is located in Yuhang District), the second production unit area (whether in Hangzhou) and pickles, Yuhang District, the relationship between sampling results, there was significance; and sampling month, sampling unit area is, the production unit area of three (if located in Zhejiang Province) and pickles, Yuhang District, the relationship between sampling results, no statistical significance.2 2001-2010 pickles in Yuhang District, sampling results to predict the model -2 log likelihood value is not big, the overall correct prediction rate of 75.85%, sensitivity and good specificity general, ROC area under the curve 0.779, model The fitting results acceptable. When the classification of 0.86, the model correctly predicted the overall rate of 51.27%, higher specificity, sensitivity is very low. But in the actual sampling work has practical significance.3 The sampling work of 2011 at Yuhang District, the region should increase the focus on small-scale production of Xiaoshan District, product sampling efforts pickles. River Town, Xiaoshan District, in particular small businesses, they should take precautions to enhance sampling in order to ensure consumer safety aspects of food pickles. Within the region in Yuhang District, should avoid sampling Chongqing Fuling mustard (Group) Co., Ltd. product, unless really necessary.4 In the long run, changes in health administration department of regulatory functions would result in Yuhang District, sampling pickles impact factors, which affect the prediction model prediction. 5 The scale of production units is too strict criteria for the classification prediction model will affect, it is difficult to predict. However, this study on the scale of production units by the lack of a development perspective, may cause deviation of forecast model construction and model prediction error is a major deficiency.