Prediction of Heart Attack and Death: Comparison Between 1 DCNN and Conventional ML Approaches

Since heart diseases influence human life significantly, so rigorous investigations and experiments are needed for the treatment of patients in the early stage. Physicians could treat after identifying the heart diseases. After getting related information about the human body, immediate prediction of heart stock is awakened by the physicians for appropriate treatment and reduce the chances of a heart attack. Again, if the probability of death due to a heart attack is detected, then the treatment may help to recover the patients. To detect heart diseases, machine learning and deep learning approaches can be coded. In this study, two datasets are experimented with conventional machine learning algorithms and deep learning algorithms. One dataset is related to the prediction of chances of heart attack and another dataset is related to the prediction of death due to heart stock. Among all approaches, the one-dimensional convolutional neural network has been shown 99.9% accuracy of prediction. More experiments on heart diseases using different machine learning approaches can conclude the appropriate model for detecting heart diseases, whereas, in this study, the one-dimensional convolutional neural network has given the best performance.