Dual-Model Machine Learning Approach for Heart Failure and Diabetes Risk Prediction

Abstract

Abstract—Health risk prediction has emerged as a significant
research domain in healthcare, particularly for chronic condi-
tions like heart failure and diabetes. This paper proposes a
dual-model machine learning framework employing Artificial
Neural Networks (ANN) and XGBoost algorithms to predict
patient risks based on clinical data. The ANN model is specif-
ically applied to heart failure prediction, while XGBoost is
utilized for diabetes prediction. Both models were trained on
benchmark datasets—UCI Heart Failure and PIMA Diabetes
datasets—achieving accuracies of 85% and 88%, respectively.
The results highlight the efficacy of combining two different
ML models to improve accuracy and reliability in healthcare
analytics.