Pacific D
Machine Learning Predictive Model on Substance Misuse in United States (306641)
*Olajide Israel Ajayi, Blue Cross NCKeywords: Machine Learning, Predictive Model
A machine learning predictive model to help government officials formulate health policies and health professionals more effectively use their time and resources combating the adverse effects of substance misuse in the United States.
Statistics on the number of lives claimed by substance misuse relative to those that died due to World War I, World War II and Vietnamese war are quite staggering. The number of emergency room (ER) visits due to substance on a daily basis, over 1,000 in the United States, further compounds the issue at stake. A pragmatic approach is required to be able to make a significant impact and stem the epidemic across the globe. Suffice it to say that individuals benefiting from the status quo based on economic perspective would fight back directly or indirectly as policies and intervention programs are initiated.
The findings from preliminary research using various models such as Classification Trees and Genetic Algorithm on prescriber data from Centers for Medicare and Medicaid Services (CMS) and other publicly available datasets at county levels in North Carolina as pilot resulted in some promising outputs and conclusions.