Preemptive Diagnosis of Hypertension Using Computational Intelligence Techniques

One of the most dangerous yet silent chronic diseases in the world is Hypertension, which has little to no symptoms and can risk strokes, heart attacks, and other sudden problems if not discovered and treated. The goal of this study is to use machine learning (ML) techniques to utilize the predictive capability of AI based systems towards early detection of this global health concern. This paper employed three different ML algorithms applied to an online dataset to develop preemptive diagnosis models for Hypertension: eXtreme Gradient Boosting (XGBoost), Histogram-based Gradient Boosting (HGB), and Light Gradient Boosting Machine (LightGBM). Dataset preprocessing included imputation of null values and normalization. Another key aspect of this study is the optimization of these models, achieved using GridSearchCV coupled with stratified five-fold cross-validation to identify suitable hyperparameters. XGBoost outperformed HGB and LightGBM, achieving the highest accuracy (92.75%) and showing enhanced predictive capabilities (precision: 92.45%, recall: 93.78%, F1 score: 93.11%). LightGBM slightly outperformed HGB with marginally better recall and F1 score. This outcome underscores the efficacy of utilizing machine learning, particularly in preemptive Hypertension diagnosis, paving the way for future research and clinical use.