International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 186 - Issue 28 |
Published: July 2024 |
Authors: Mawuli Agboklu, Benjamin Lartey, Frederick Adrah, Dennis Lajeunesse |
![]() |
Mawuli Agboklu, Benjamin Lartey, Frederick Adrah, Dennis Lajeunesse . Towards Proactive Heart Health: A Machine Learning - Powered Approach for Chronic Heart Failure Detection. International Journal of Computer Applications. 186, 28 (July 2024), 30-35. DOI=10.5120/ijca2024923780
@article{ 10.5120/ijca2024923780, author = { Mawuli Agboklu,Benjamin Lartey,Frederick Adrah,Dennis Lajeunesse }, title = { Towards Proactive Heart Health: A Machine Learning - Powered Approach for Chronic Heart Failure Detection }, journal = { International Journal of Computer Applications }, year = { 2024 }, volume = { 186 }, number = { 28 }, pages = { 30-35 }, doi = { 10.5120/ijca2024923780 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2024 %A Mawuli Agboklu %A Benjamin Lartey %A Frederick Adrah %A Dennis Lajeunesse %T Towards Proactive Heart Health: A Machine Learning - Powered Approach for Chronic Heart Failure Detection%T %J International Journal of Computer Applications %V 186 %N 28 %P 30-35 %R 10.5120/ijca2024923780 %I Foundation of Computer Science (FCS), NY, USA
Myocardial infarction, more commonly known as “heart attack” is one of the most dangerous diseases worldwide. Timely detection and intervention are crucial for saving the lives of patients and reducing mortality rates. Beside traditional clinical interventions, machine learning (ML) techniques have garnered considerable attention for their potential in aiding the early detection of heart disease in recent years. In this study, we will use ML algorithms such as Random Forum (RF), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost) and Logistic regression (LR) to develop models to predict the possibility of chronic heart failure in patients hospitalized with myocardial infarction 72 hours after their hospitalization. Varied optimization techniques were applied to these models to improve their predictive outcomes. The models were evaluated using metrics such as accuracy, precision, recall, f1, mcc and confusion matrix and compared against each other to determine which of them generated better results. The XGBoost algorithm demonstrated superior performance compared to the other models. The dataset was collected from UCI machine learning repository with the database containing 1700 patient records and 111 input features.
No references available