Heart disease classification system using optimised fuzzy rule based algorithm

Abstract

Heart disease prediction and identification is a difficult task which needs much experience and knowledge. In order to predict the heart disease, we introduce a technique named as RBFL prediction algorithm. The overall process of the RBFL prediction algorithm is divided into two main steps, such as 1) feature reduction using LPP algorithm, and 2) Heart disease classification by means of rule based fuzzy classifier. Initially, LPP algorithm is employed to recognise the related attributes and then fuzzy rules are produced from the FFBAT algorithm. Next, the fuzzy system is designed with the help of designed fuzzy rules and membership functions so that classification can be carried out within the fuzzy system designed. At last, the experimentation is performed by means of publicly available UCI datasets, i.e., Cleveland, Hungarian, Switzerland datasets. The experimentation result proves that the RBFL prediction algorithm outperformed the existing approach by attaining the accuracy of 76.51%.