A Comparative Assessment Study on Machine Learning Classifiers for Cardiac Arrest Diagnosis and Prediction

Abstract

Heart attack, also known as cardiac arrest, encompasses various heart-related disorders and has been the leading cause of death worldwide in recent decades. Many risk factors are linked to heart illness, and there is a pressing need for accurate, effective, and practical methods to make an early diagnosis and treat the disease. In order to appropriately categorise and predict heart attack patients with minimal features, this study tested alternative algorithms for classification of the dataset. An in-depth comparison is made using pre-processing and standardisation techniques on the UCI dataset, and ensemble algorithms over supervised algorithms, as well as comparing custom neural net design to pre-defined procedures. With the total of 17 used so far, Random Forest (RF) gives a maximum accuracy of 96.5%, which is examined from the survey work. Future study could combine several machine learning techniques to produce a more comprehensive model, which could help health care practitioners make better judgments.