Predicting Diabetes Using Machine Learning : XGBoost and the PIMA Dataset

Diabetes is a significant health issue that needs innovative approaches to be managed and detected early. Ma-chine Learning (ML) has demonstrated a significant potential in improving its prediction and diagnosis. The present paper highlights a comprehensive comparison of different machine learning algorithms (eXtreme Gradient Boosting (XGBoost), Random Forest (RF), K-Nearest Neighbor (KNN) and C-Support Vector Classification (SVC)) applied to the Pima Indian dataset, a well known resource in the diabetes research. After rigorous experimentation and evaluation, we have identified the most effective ML algorithm that surpasses existing models in accuracy, precision and recall. This work employs a supervised ML approach, XGBoost, that achieved the best accuracy after exploring its parameters.