The Prediction And Control Of Food Safety Based On Bayesian Statistics

Food safety has bee referred to as the global problem of growing concern at present.Food enterprises established a food traceability system to collect index data that needed to becontrolled. And these index data which could bring potential risks that existed in everyproduction progress.We searched for reasons when the food safety occurred before, but wecan not prevent the problem in the future. The method we have to solve is to find out factorsthat impact food safety, then control on this basis.This article first defines the potential risks of the food and divides the level of risk, thenswitches the raw data to the data that represented by the potential level of risk. Determinecritical control point of every process according to Hazard Analysis and Critical Control Point（HACCP） system. Get weight of every control point by Analytic Hierarchy Process.Established a comprehensive evaluation model of food safety risk which was impacted by thepotential risks of the various aspects. Further more, take potential risk level of every aspect asa priori information of potential risk random variables and sampling data as the sample. Then,take advantage of the Bayesian Statistics to get posterior distribution of the variable basedon priori information and sample information and take Maximum Likelihood Estimation ofthe posterior distribution as discriminant of fanal potential risk degree. To reason therelationship of potential risk、the relationship of food safety risk and potential risk,weextracted association rules of the potential risk and the rules of the change that food safetybaseed on pontential risk of every control impacts. Predict food sefety risk based on TemporalAssociation Rules. Diagnose and control food safety risk that may arise.