Research on food safety prediction method based on k-means clustering algorithm

Abstract

Aiming at the problem of food risk prediction, this paper proposes a method based on clustering algorithm to predict product risk by raw material risk. Firstly, based on the provincial supply chain closed-loop hypothesis, this paper proposes the selection method of clustering indexes for products and their raw materials. Secondly, this paper uses the k-means clustering algorithm to cluster the products and the corresponding raw materials respectively, then based on the clustering Class results automatically determine the high-risk categories of the products and their raw materials. Finally, the analysis of the experimental data of the 8 categories of products and their raw materials shows that the ratio of the high-risk categories of products and the ratios of the corresponding high-risk categories of raw materials have a strong positive correlation. The experimental results prove the rationality of the raw material clustering index selection method proposed in this paper and the correctness of the method of predicting product risk based on the raw material risk based on the clustering algorithm.

Keywords

• food safety
• food raw materials
• k-means clustering
• pearson correlation coefficient