Zhiyuan Song

SL4HOI: Supervised Learning for Predicting Health Opportunity Index across States

Authors: Zhiyuan Song, et al.

Published in: Knowledge Discovery and Data Mining (KDD) (2024)

Abstract: Health disparity is a critical issue, with access to health opportunities unevenly distributed across populations. The Virginia Health Opportunity Index (HOI) provides a comprehensive measure of the social determinants of health (SDH), yet its complex computation limits its applicability beyond Virginia. This research aims to simplify the HOI calculation process and extend its utility to other states by developing a Supervised Learning (SL) model using readily available data from the American Community Survey (ACS). We acquire and process ACS data for Virginia, train and validate a Random Forest model to predict HOI, and test its applicability in North Carolina and California. Our model demonstrates robust performance, with positive correlations between predicted HOI and life expectancy and low p-value in all states tested. This study has implications for public health policy, enabling more accessible and generalizable tools for assessing health opportunities and facilitating targeted interventions to promote health equity.