The Applied Research of Dynamic Clustering Algorithm in Identifying Vegetable Oil Species

A more practical, efficient, fast identification for food raw materials is favorable to improve the current food security situation. In order to improve this kind of condition, this paper presents a vegetable oils discrimination based on improved K-Means algorithm and according the GC of vegetable oil. And this algorithm is improved in selecting original center of clustering so that the traditional K-Means algorithm can get a global optimal solution instead of a local optimal one, and get a more stable clustering result. The experiment result shows that the algorithm gains an evident improved effect and a good performance.