Transformations  in
Business & Economics

SOCIO-ECONOMICS OF OPIOIDS CRISIS BASED ON THE GRADIENT BOOSTING DECISION TREE MODEL: A CASE FOR SOME STATES OF THE US³
Jinming Zhou, Weihua Su, Valdemaras Makutenas

ABSTRACT. There are implications for important sectors of the U.S. economy as well. For example, if the opioid crisis spreads to all cross-sections of the U.S. population (including the college-educated and those with advanced degrees), businesses requiring precision labor skills, high technology component assembly, and sensitive trust or security relationships with clients and customers might have difficulty filling these positions. Further, if the percentage of people with opioid addiction increases within the elderly, health care costs and assisted living facility staffing will also be affected. Based on the Gradient Boosting Decision Tree (GBDT) model, this paper analyzes the correlation analysis with support vector machine algorithm, geographic information marker and so on, and puts forward some relevant strategy suggestions. Firstly, the parameters of the model are constantly adjusted so that the model adapts to the parameters, and the accuracy of the model is finally 0.8438. Secondly, combined with analysis and related literature, this paper enumerates the reasons for the growth of opioid use and addiction in three points and the reasons for two growth. The number of different drugs is calculated according to the model, and the threshold for drug identification in each county reached 5501 of the county's total drug identification, and Philadelphia was the most likely to exceed the drug identification threshold level. With the support vector machine algorithm to process the data, it is concluded that the function of the support vector machine regression model is proposed. Finally, three strategies are proposed against the opioid crisis. The remaining variables of this model still have an effect on synthetic opioids. In order to make the results universal, this article fills in the remaining variables and calculates the corresponding amount of synthetic opioids. The results show that the propagation characteristics of the use of synthetic opioids and heroin events are increasing with the increase of time. And the use of multiple regions of opioid drugs has a certain radiation influence on the surrounding area which has a piece-wise structure. It can be concluded from the characteristic graph that the problem of aging in the social economic structure has the greatest influence on synthetic opioids. And the problem of social aging has a certain impact on the opioid crisis .

KEYWORDS:  Gradient Boosting Decision Tree (GBDT), Support Vector Machine (SVM), Opioids, GIS, prediction.

JEL classification:  I18, C61, G51.

³Acknowledgments:  The authors acknowledge the International Narcotics Control Board for making data available. The work was supported by National Social Science Fund General Project (No. 20BTJ048), Support Program for Outstanding Young Talents in Colleges and Universities of Anhui Province (gxyq2020032).
Contributors. J.M. Zhou designed the study; collected, analyzed, and interpreted the data; and drafted the article. J.M. Zhou contributed to the design of the work, analyzed the data, and contributed to drafting the article. W.H. Su contributed to the design of the work and critically reviewed and revised the article. All authors approved the final article as submitted.
Conflicts of Interest. We declare no conflicts of interest.