A Comparative Study of Machine Learning Algorithms for Early Detection Of Heart Disease
Cardiovascular diseases remain a leading global cause of mortality, making early and accurate detection critical for effective medical intervention. Traditional diagnostic methodologies frequently rely on invasive procedures or late-stage symptom recognition, which can significantly delay treatment. To address this challenge, this paper presents a comprehensive comparative analysis of machine learning algorithms designed to facilitate the non invasive, early detection of heart disease. The study systematically evaluates the predictive capabilities of Logistic Regression, Support Vector Machines, Decision Trees, Random Forest, and Grid Random Forest models applied to publicly available clinical datasets encompassing 11 distinct physiological features, including Age, Sex, Chest Pain Type, Resting BP, Cholesterol, Fasting BS, Resting ECG, Max Heart Rate, Exercise Angina, OldPeak, and ST_Slope. A core focus of this research is the vital role of data preprocessing, robust feature engineering, and hyperparameter optimization. Model performance is rigorously assessed through cross-validation techniques, employing standard evaluation metrics including accuracy and F1-score.
Authors
- Hariom Pawar
- Aryan Harak
- Pratik Haral
- Parth Harde
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24415955
- Primary Topic
- Artificial Intelligence in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 0.00