VitalScore: Heart Disease Risk Classification Using Bayesian Logistic Regression and Support Vector Machines
IARJSET · 11.11.2026 · 13 (11)
DOI: 10.17148/iarjset.2026.131144 ↗Kurz gefasst
Fachartikel in der Zeitschrift IARJSET, online erschienen am 11.11.2026. Verfasst von S. Lakshman und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Heart disease risk models commonly return a point estimate without indicating how uncertain that estimate is for an individual.This paper presents VitalScore, a web-based decision-support prototype that combines Bayesian logistic regression with an RBF-kernel support vector machine (SVM) baseline.The models were evaluated on 918 records from the public Heart Failure Prediction dataset using a stratified 80/20 split.Preprocessing treated zero-coded cholesterol and resting blood-pressure values as missing, encoded nominal variables using one-hot indicators, and standardized features using training data only.On the held-out set of 184 patients, Bayesian logistic regression achieved 90.2% accuracy, 90.4% precision, 92.2% recall, an F1-score of 0.913, and an ROC-AUC of 0.935.The SVM achieved 87.0% accuracy and an ROC-AUC of 0.933.The Bayesian model additionally produced patient-level 95% credible intervals and additive feature contributions.When ST slope was unrecorded, predictions were averaged across its categories.These results show the feasibility of presenting risk, uncertainty, and interpretable contributions in one application; they do not establish clinical effectiveness.
Zitieren
APA
Lakshman, S., B, B. C., & Poornima, S. (2026). VitalScore: Heart Disease Risk Classification Using Bayesian Logistic Regression and Support Vector Machines. IARJSET, 13(11). https://doi.org/10.17148/iarjset.2026.131144
Vancouver
Lakshman S, B BC, Poornima S. VitalScore: Heart Disease Risk Classification Using Bayesian Logistic Regression and Support Vector Machines. IARJSET. 2026;13(11). doi:10.17148/iarjset.2026.131144
BibTeX
@article{Lakshman2026,
author = {Lakshman, S. and B, Banshika Chowdhury and Poornima, S.},
title = {VitalScore: Heart Disease Risk Classification Using Bayesian Logistic Regression and Support Vector Machines},
journal = {IARJSET},
year = {2026},
volume = {13},
number = {11},
doi = {10.17148/iarjset.2026.131144}
}
EUVIMED