A hybrid ARIMA-ANN model for Internet Food safety risks

As Internet technology becomes more and more mature, Internet food sales have become more and more problematic. problems from raw materials to food processing to food distribution are not easily regulated because of online sales. Firstly, crawl the take-out sales data of cities across the country, based on the data the time series prediction model is established to realize the prediction of business and regional risk values, From this obtains the merchant early warning as well as reminds the consumer's function. According to the data analysis, the ARIMA+ANN model is used for short-term prediction. Then the data analysis results of different regions are compared, and the data analysis of both sides of time and space is conducted, The value at risk ranking allows consumers to make a clear comparison of effects, In this way, we can realize the real and effective supervision of store owners from the economic level, realize the real and effective supervision of Internet food sales, and ensure food safety.