Performance Analysis of Rule Based Algorithms Applied to a Cardiovascular Dataset

Abstract

Cardiovascular disease is one of the dominant concerns of society, affecting millions of people each year. Early and accurate diagnosis of risk of heart disease is one of major areas of medical research, aimed to aid in its prevention and treatment. Most of the approaches used to predict the occurrence of heart disease use single data mining techniques. However, performances of predictive methods have recently increased upon research into hybrid and alternative methods. This paper analyses the performance of logistic regression, support vector machine, and decision trees along with rule-based hybrids of the three in an attempt to create a more accurate predictive model.