Performance Evaluation of Regression-Based Machine Learning Algorithms for Myocardial Infarction Prediction

Myocardial Infarction (MI) is a disease condition where the supply of blood to the heart or some parts of the heart is obstructed as a result of an occluded coronary artery. Myocardial Infarction is very common in Europe and the United States and is arguably one of the leading causes of death in those countries. Timely detection of MI can reduce the cost of treatment but early prediction can be very helpful in preventing the development of MI in the first instance. Supervised Machine learning even in the presence of uncertainties can be used to make predictions based on trained models from some known inputs and/or outputs. In this paper, the performance of different regression-based Machine Learning algorithms has been carried out using MATLAB. The performance analysis was done in terms of the ability of those algorithms to predict future response events based on changes from predictors present in a dataset. Root Mean Square Error (RMSE), R-Squared, Mean Squared Error (MSE), and Mean Absolute Error (MAE) are four important metrics utilized in the performance analysis. It was discovered from the analysis that ordinary linear regression trained model outperformed other regression based models with respect to the four metrics mentioned above.

کلیدواژه ها:

Myocardial Infarction (MI) ، Supervised Machine Learning ، Regression Learning ، electrocardiograph signals