Author(s)

Xinmiao Wang

Corresponding Author

Xinmiao Wang

Abstract

On the basis of clustering analysis, the prediction of future drug reports. For the first part, we established a propagation feature model of the reported synthetic opioid and heroin cases. For each county, the clusters of each state are packaged by cluster analysis, and the clustering centers of the districts of the counties in each state are given. For the second part, we established the BP neural network model, and the feasibility test of the neural network trained with the 2010 data was conducted with the 2011 data, and it was found that the model had a high credibility. Finally, take 2017 as an example, the model gives the forecast of 2018 drug reports and the counties that need to worry.

Keywords

Cluster analysis, BP neural network