AI-Driven Prediction and Hierarchical Classification of Heart Disease Severity

Heart disease poses a significant threat to human existence because of its high mortality and morbidity rates. For early treatment, localization, and countermeasure, precise anticipation and conclusion are becoming increasingly important. The Internet of Things and artificial intelligence make it easier for doctors to find, evaluate, and diagnose cardiac diseases. However, the majority of prediction models only assess the severity of an individual's illness and only predict whether they are ill or not. In this study, we present a hypothesis model for simultaneous assumption for both matched and distinct request coronary illness that is based on machine learning (ML). We first developed a fuzzy-GBDT strategy that combines cushy reasoning with gradient boosting decision tree (GBDT) to advance the matched portrayal assumption and reduce data complexity. To avoid overfitting, the pressing is completed with GBDT. The severity of cardiac disease is also included in the multi classification assumption based on Bagging-Fuzzy-GBDT. The assessment shows that the Packing Fluffy GBDT is incredibly dependable and exact for both twofold and different get-together presumptions.