Intelligent Heart Disease Diagnosis: Analyzing Predictive Accuracy of Machine Learning Models

Abstract:-Heart disease remains a leading cause of mortality worldwide, necessitating early diagnosis and preventive healthcare strategies. With the increasing integration of machine learning (ML) into medical analytics, predictive models have become instrumental in enhancing the accuracy and efficiency of heart disease diagnosis. This review investigates recent advancements in heart disease prediction using ML algorithms,Emphasis is placed on widely adopted models such as Decision Trees, Random Forest, Naive Bayes, Logistic Regression, K-Nearest Neighbors, and XGBoost.These algorithms have demonstrated significant potential in identifying complex patterns in clinical datasets, surpassing traditional statistical methods in adaptability and predictive power. Various studies have highlighted the strengths of each algorithm: Decision Trees and Logistic Regression offer high interpretability, while ensemble techniques like Random Forest and XGBoost deliver superior accuracy and robustness. Naive Bayes proves effective with limited data, and KNN is noted for its performance in normalized, noise-free environments.The review also discusses the relevance of datasets like the UCI Cleveland Heart Disease Dataset and the critical role of preprocessing techniques such as normalization, imputation, and feature selection. Evaluation metrics including accuracy, precision, recall, F1-score, and ROC-AUC are analyzed for their effectiveness in assessing diagnostic performance. Furthermore, hybrid and ensemble methods have shown promise in boosting predictive outcomes through model integration and optimization.This paper concludes by emphasizing the importance of algorithm selection, data quality, and preprocessing in developing reliable ML-based heart disease prediction systems. The insights presented aim to guide future research and support clinical decision-making through intelligent, data-driven solutions.

---