부산대학교 도서관

The genesis of cardiovascular disease is still a global issue that has not been addressed, and the high suffering, impairment, and death rates that are associated with cardiovascular illnesses are the disease's primary features. Therefore, there is a need for artificial intelligence (AI) tools that are both effective and quick in the earlier detection of potential results in individuals who have cardiovascular disease. The Internet of Things (IoT) is growing more pervasive, and this is helping to improve the possibilities of AI technologies. Sensors connected to the internet of things are used to gather data, which is then retrieved and forecasted using technology to predict. Common machine learning technologies that are currently in use are not very good in handling data disparities and have a rather poor level of model accuracy rate. The findings of this article propose a classification algorithm aggregation approach that relies on stackable prototype merging to address this problem. These authors take into account the information will be analyzed and training methodologies used by various algorithms. In order to prevent fitting problem, we utilize a basic linear classifier known as Logistic Regression (LR) as that of the macro classifier. We verified the methodology by utilizing a fused Heart Dataset that was compiled from numerous machine learning libraries at the University of California, Irvine, as well as another Heart Attack Dataset that was made publically accessible, and we compared it to 10 single classifier models. According to the findings of the experiments, the stacking classifier that was developed is superior to other classifiers in terms of both its accuracy and its application.