Predicting Hypertension Amongst Black Women Using Machine Learning Models
Author(s)
Klasra, Ahmed Rauf
Advisor(s)
Editor(s)
Collections
Supplementary to:
Permanent Link
Abstract
Black women in the United States are disproportionately affected by adverse maternal and
general health outcomes related to high blood pressure and hypertension. To investigate and mitigate this issue, we trained multiple machine learning models to predict hypertension amongst black women while using SHAP analysis to identify predictors. We found Gradient Boosting Machines the best model for our use case. SHAP analysis revealed features such as BMI, weight, and smoking found to be common predictors of hypertension amongst black women. Novel features identified by the analysis included widowed black women being at higher risk of developing hypertension and immigrant black women at lower risk of developing hypertension
Sponsor
Date
Extent
Resource Type
Text
Resource Subtype
Undergraduate Research Option Thesis