How to evaluate the safety risks of agricultural products is a crucial issue in current China. Based on agricultural products source, we first identify key elements to establish an evaluation framework with 5 first grade indicators and 17 secondary indicators. Subsequently we apply back-propagation model to set up a neural network risks evaluation model and train the established neural network with data collected from main Chinese food bases. Simulation shows the maximum error between trained neural network and the evaluation score of the experts is rather small and certify that back-propagation model can be applied to evaluate the safety risks of agriculture products.
展开▼