Study And Construction On Food Safety Traceability Assistance System Based On Machine Learning

Food safety has attracted a lot of attentions,because it has huge impacts on human health and social stability.Government departments and related agencies try to improve the food safety aspects of policy and regulations,while consumers pay attention to food safety because of their own healthy problem.As the food sampling programs are implemented time to time,in local areas or throughout the country,in order to guarantee food safety and improve food quality,a huge amount of data has been accumulated.The hidden patterns in the data and their potential values are drawing more and more attentions from the research community,through appropriate algorithms and analysis can predict food quality and trace the source of food problems.Extreme learning machine(ELM),a method based on neural network,is employed in this paper to mine the relationship between food quality and food category,manufacturing site and season,inspection site and season,and many other attributes,and to predict food quality with a certain accuracy.In this paper,the food sampling inspection data,through the ELM algorithm for classification prediction of food quality,and can trace all the information on food traceability process and correlation analysis.ELM is applied to the Food Safety Traceability Assistance System for the first time in this paper,and achieved good results.At the same time,according to the characteristics of the ELM,this system can provide a better structure for future sampling plan.Experimental results on real data sets indicate that the ELM approach has better prediction precision and generalization ability than logistic regression that was adopted in preceding work.The patterns obtained are helpful in making more effective food sampling plans and for more targeted food safety tracing.