Author(s)

Xudong Yang

Corresponding Author

Xudong Yang

Abstract

Opioid crisis has contributed to loads of damages in both life expectancy and economical development in U.S. In order to tackle this extremely urgent issue, we need clear purposes and direct efforts, which means the necessity to uncover influential factors.In this paper, we are encouraged to build suitable mathematical models with loads of statistical data provided and required processed, which includes drug reported amount and some economical indexes of five state. First, we establish a linear propagation model and get three kinds of equation: county-county, county-state and state-state, and we evaluated the possibility of some counties as sources from the county-state weight matrix. Then we use the value of the weight matrix to extract the main influencing features and perform a simple analysis.Finally, we come up with the conclusion including the possible start of pointed opioid use, and the key feature of addicted people.

Keywords

Data Processing, Linear Model, Feature Extraction