In today's era cardiac arrest and other heart diseases are the most common problem in majority of people, and there are various factors that act as backbone of this problem like people are not paying attention towards health mainly because of work stress, laziness, substandard quality of food that results in increasing cholesterol and untimely diagnosis of heart disease due to lack of technology, methods used in diagnosing these diseases and consequently having a lot of tests. A lot of research and medical supporting systems are developing day by day, however, every system have its various features or advantages and limitations which are unknown to either side. This paper aims to study different machine learning algorithms on dataset to predict the possibility of a cardiac arrest based on various controlled and uncontrolled variables.