Machine Learning-Based Heart Disease Classification for Symptom-Driven Diagnostics

Aug 5, 2026

Upload by Ezra

Heart disease prevalence continues to rise, and early identification remains difficult for clinicians due to limitations in conventional diagnostic paradigms. This study applies machine learning methods to improve symptom-driven early diagnosis by preparing a dataset with demographic and clinical characteristics, performing normalization, removing outliers, and using principal components analysis to optimize feature dimensions. Multiple supervised classifiers are evaluated using confusion matrices, accuracy, and ROC AUC, with Random Forest delivering the strongest validation results and high discriminative performance, supporting faster, more efficient screening that can help reduce mortality.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages21

Reading Time53 min

AccessFree to download

Introduction

Heart disease and cardiovascular risk factors

Motivation for AI-driven early detection

Methodology

Dataset preparation and feature engineering

Dimension reduction with PCA

Model Training and Evaluation

Supervised classifiers and metrics

Results and validation

Conclusion

Machine Learning-Based Heart Disease Classification for Symptom-Driven Diagnostics Muhammad Talha Jahangir*1, Tahir Abbas2, Muhammad Hamza Khan3, Amjad Ali4, Burhan Mughees5, Afaq Ahmad6, Muhammad Ahsan Jamil7 1Department of Computer Science, MNS-University of Engineering and Technology, Multan, Pakistan. 2Department of Computer Science, TIMES Institute, Multan, Pakistan. 3Department of Electrical Engineering, MNS UET, Multan, Pakistan. 4Department of Information Technology, Bahauddin Zakariya University, Multan, Pakistan. 5FAST-National University of Computer and Emerging Sciences, Faisalabad Campus 6Punjab Tianjin University of Technology, Lahore, Pakistan 7Institute of Computing, MNS-University of Agriculture, Multan *[Correspondence:](Correspondence: mtalhajahangir@mnsuet.edu.pk)[ ](Correspondence: mtalhajahangir@mnsuet.edu.pk)[mtalhajahangir@mnsuet.edu.pk](Correspondence: mtalhajahangir@mnsuet.edu.pk) Citation | Jahangir. M. T, Khan. M. H, Ali. A, Mughees. B, Ahmad. A, Jamil. M. A,“Machine Learning-Based Heart Disease Classification for Symptom-Driven Diagnostics”, IJIST, Vol. 6 Issue. 4 pp 1768-1788, Oct 2024 Received| Oct 02, 2024 Revised| Oct 22, 2024 Accepted| Oct 25, 2024 Published| Oct 26, 2024. Heart diseases are increasing over the period while identifying cardiac diseases at an early stage continues to pose a challenge. This study focuses on the application of AI specifically in machine learning to improve early diagnosis of this ailment. We overcome the limitations of conventional diagnostic paradigms. Normalization was performed on a dataset with demographic and clinical characteristics data, outliers were removed, and principal components analysis was used to enhance and decrease dimensions to get optimized results. Supervised learning classifiers such as Support Vector Machine, Decision Trees, Random Forests, Logistic Regression, K- Nearest Neighbors, and Naive Bayes evaluated based on metrics such as confusion matrix, accuracy, and ROC AUC scores. Of all the models created, the Random Forest model was found to have the best internal validation results with an accuracy of 1.0 as well as test and training ROC AUCs of 0.97 for detecting heart disease cases and noncases. It is evident that developing an AI model for the diagnosis of heart disease provides promising results of faster and more efficient diagnosis reducing the mortality rates of the disease. Keywords: Heart Disease, Machine Learning, Classification, Random Forest Classifier, KNearest Neighbor (KNN), Support Vector Machines (SVM), PCA. Introduction: The heart is one of the most vital organs in our body, with its primary function being to move blood throughout the body. Despite this, there are several health issues that it might cause. Problems with the heart and blood vessels are known as cardiovascular diseases (CVD). The most common killer in the modern world is cardiovascular disease, which manifests mostly as heart attacks and strokes. About 32% of all fatalities occur because of cardiovascular disease, which accounts for 17.9 million deaths annually [1] . Four out of five deaths caused by cardiovascular disease occur because of heart attacks or strokes [2] . Because of its size and importance, the heart requires special attention. Predicting cardiac problems is crucial, necessitating comparative study in this area, as most diseases have some connection to the heart. More effective disease prediction algorithms are required because most people die because their illnesses are only discovered at a late stage because of inaccurate medical instruments [3] . There are a lot of risk factors for heart disease, such as being overweight, having high cholesterol, smoking, not eating well, having diabetes, and having irregular heart rhythms [4] . Problems with the heart's melody, valves, tissues, infections, blood arteries, or congenital abnormalities are all examples of heart diseases. Globally, coronary disease (CHD) ranks

FAQs

What is the main goal of the study?

To apply machine learning for improving early, symptom-driven diagnosis of heart disease and address limitations of conventional diagnostic approaches.

How was the dataset processed before model training?

Normalization was performed, outliers were removed, and principal components analysis (PCA) was used to reduce and optimize feature dimensions.

Which classifier performed best according to the reported results?

Random Forest showed the best internal validation performance, with accuracy reported as 1.0 and ROC AUC values of 0.97 for detecting both heart disease cases and non-cases.

Aug 5, 2026

Upload by Ezra