A Novel Hybrid Binary Farmland Fertility Algorithm with Naïve Bayes for Diagnosis of Heart Disease

Abstract

One of the essential aims of intelligent algorithms concerning the diagnosis of heart disease is to achieve accurate results and discover valuable patterns. This paper proposes a new hybrid model based on Binary Farmland Fertility Algorithm (BFFA) and Naïve Bayes (NB) to diagnose heart disease. The BFFA is used for Feature Selection (FS) and the NB for data classification. FS can be employed to discover the most beneficial features. Four valid and universal UCI datasets (Heart, Cleveland, Hungary and Switzerland) were used to diagnose heart disease. Each dataset included 13 main features. The evaluation of the proposed model is simulated in MATLAB 2017b. The number of features in four datasets of Heart, Cleveland, Hungary and Switzerland is equal to 13, which was reduced to six for each dataset through the BFFA to better the efficiency of the proposed model. For evaluation, the accuracy criterion, the criterion of accuracy in the proposed model for all features in the four datasets Heart, Cleveland, Hungary and Switzerland, is equal to 82.25%, 86.91%, and 89.32% 89.24%, respectively. Results of the proposed model showed appropriateness in comparison to some other methods. In this paper, the proposed model was compared with other methods, and it was found out that the proposed model possessed a better accuracy percentage.