Predictive Analysis of Heart Disease Using Machine and Deep Learning Approaches

Heart disease remains one of the leading causes of mortality worldwide, making its early detection critically important. This seminar report explores recent advancements in the prediction of heart disease using machine learning (ML) and deep learning (DL) approaches, based on five contemporary research studies. The comparative analysis covers various supervised ML algorithms including Decision Trees, Random Forest, Support Vector Machines (SVM), K-Nearest Neighbours (KNN), Logistic Regression, AdaBoost, and ensemble models, as well as deep learning models such as neural networks. Most of the studies utilized datasets like the Cleveland Heart Disease dataset from the UCI repository. Techniques such as data preprocessing, feature selection (using PCA, Isolation Forest, and other filters), and dimensionality reduction were integral in improving prediction accuracy. Accuracy across models varied from 87% to 94.2%, with hybrid models and deep learning frameworks showing the most promise in terms of precision and generalization. This report highlights the potential of AI-driven models in reducing diagnostic errors, optimizing treatment strategies, and ultimately enhancing clinical outcomes in cardiovascular healthcare.