Robust Framework for Detecting Heart Stroke using Advanced Machine Learning Techniques

Strokes of the heart are responsible for an increasing number of deaths that occur on a daily basis, and it is anticipated that this pattern will continue for the foreseeable future. It is unfortunate that determining the presence of such illnesses in human people is a challenging and time-consuming task. The management of tasks that are of such a complex nature might benefit from the utilization of data sets as a valuable tool. The only way to effectively predict the occurrence of heart strokes is through the utilization of automation. This is due to the fact that it is essential to continue monitoring the heart rate. The arrhythmia dataset from the Massachusetts Institute of Technology (MIT) is utilized in this particular piece because it is one of the datasets that is beneficial to us. There are a variety of data mining methodologies that can be utilized in order to accomplish the goal of automating the task that was defined. Within the scope of this article, the approaches that were employed encompass decision trees, naïve bayes, the kNN, and the Random Forest techniques. Therefore, the objective of this research is to construct a comparison between the algorithms that have been explained above and to evaluate which one is more accurate in terms of finishing the assignment. We are able to claim that the algorithm Random Forest has achieved a level of accuracy that is $\mathbf{9 9 \%}$, which is the highest level of accuracy among all of the algorithms. This is the conclusion that we are able to reach at the end of the day, after all of the assessments have been completed.