ABSTRACT
Cardiovascular disease is the serious and utmost occurring critical and prevalent disease that need immediate cure and preventive action. A survey having data from more than 190 countries, estimates around 17.3 million deaths due to heart disease, which expects to cross 23.6 million by 2030 (American Heart Association statistical report tracks global figures for first time n.d.). Machine learning with its immense contribution helped both specialized and non-specialized doctors in diagnosing and taking corrective action for prevention of this crucial disease. Many researchers experimented with several traditional and hybrid models bring down the heart disease death rates. The traditional approaches are not sufficient to predict the heart disease, the machine learning based hybrid systems such as fuzzy logic with neural networks, support vector machine and genetic algorithm have emerged as better solution and outperformed as compared to traditional ones. This paper proposes two major objectives - the first objective describes the comparative analysis of traditional and machine learning based hybrid methodologies along with their concern algorithmic techniques, technical advantages, which are signifying the best-optimised procedures used for decision-making; the second objective describes the past designed ML based methodologies along with accuracy, sensitivity and specificity implicated outcomes. Theses machine learning based decisions could help in predicting the disease and helping doctors in recommending medical prescription. This study included papers which are published after 2016.
