Abstract :

Heart Failure Remains One Of The Leading Causes Of Hospital Admission And Death Across The World, And Catching It Early Is Often The Difference Between A Manageable Condition And A Medical Emergency. This Project Looks At Whether Routinely Recorded Clinical Measurements - The Kind Already Collected During A Normal Patient Check-up - Can Be Used To Build A Machine Learning Model That Flags Patients Who Are Likely To Be At Risk. Before Any Model Was Trained, The Clinical Records Were Cleaned Up And The Numeric Attributes Were Brought Onto A Common Scale, Since Several Of The Algorithms Used Later Are Sensitive To Features That Differ Widely In Magnitude. Five Supervised Algorithms Were Trained And Compared On The Same Processed Data - Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, And An Optimized Gradient Boosting Model (XGBoost) - And Judged Using Accuracy, Precision, Recall, F1-score And The ROC-AUC Curve, Rather Than Accuracy On Its Own. Among The Evaluated Models, Random Forest Achieved The Highest Classification Accuracy (83.33%) And The Highest ROC-AUC (0.892). Logistic Regression (81.67%) And XGBoost (81.67%) Also Performed Competitively, While Decision Tree And Support Vector Machine Produced Comparatively Lower Performance. Taken Together, The Results Suggest That A Well-prepared Dataset Paired With A Sensibly Tuned Model Can Act As A Useful Second Opinion For Clinicians Screening For Heart Failure Risk - Not As A Replacement For Their Judgement, But As An Extra Signal That Can Help Prioritize Which Patients Need Closer Attention. It Is A Small-scale Study, But It Points To The Practical Value That Even Modest, Well-executed Machine Learning Can Add To Everyday Clinical Decision-making.

Published:

18-8-2026

Issue:

Vol. 26 No. 8 (2026)

Page Nos:

986-991

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite