Comprehensive machine learning models for effective Cardiovascular Disease prediction
|
Full Text |
Pdf
|
|
Author |
Vinayagam P., Gunasundari C., Kumaran C. and Anne Jenefer F.
|
|
e-ISSN |
1819-6608 |
|
On Pages
|
1721-1730
|
|
Volume No. |
20
|
|
Issue No. |
19
|
|
Issue Date |
January 10, 2026
|
|
DOI |
https://doi.org/10.59018/1025194
|
|
Keywords |
CVD, ML models, data processing techniques, RF, KNN, and NB.
|
Abstract
Every year, at least more than 2,200 people die per day due to Cardiovascular Disease (CVD), which is commonly caused due to high blood pressure, stress, smoking, and reduced physical activity. However, accurate diagnosis of CVD is generally a challenging task; thus, this paper proposes Machine Learning (ML) based prediction methods, which are rapidly increasing, as they are capable of recognizing complex patterns from data. Additionally, to ensure higher quality output of ML models, it is essential to pre-process data; thus, various data processing techniques involving data cleaning, data exploration and analysis, data splitting, and data scaling are integrated to improve the prediction accuracy by removing missing or inconsistent data. Furthermore, three ML based classification techniques, such as Random Forest (RF), K-Nearest Neighbor (KNN), and Naïve Bayes (NB), are evaluated in this paper to determine the finest classifier model. The proposed approach for predicting CVD is evaluated using a dataset in Python software, and the obtained result shows that RF considerably attained an accuracy of 98.54%, while KNN attained 83.41% and NB attained 80%. Hence, the proposed technique provides a great impact for healthcare by accurately predicting CVD with higher efficiency.
Back