Enhancing the precision and accuracy of renal failure diagnosis using the modified support vector machine algorithm and dragonfly algorithm

Various methods have been proposed to diagnose renal failure based on data mining and artificial intelligence techniques. The use of data mining approaches, renal failure could be predicted based on several features and risk factors that exacerbate the condition. At the same time, each is associated with issues such as complex computation and a long implementation period. Moreover, these methods have varied accuracies due to their dependency on algorithms, performance, and nature of data. The present study aimed to propose an approach to increasing the accuracy and efficiency of the renal failure diagnosis. To this end, we developed a feature selection method based on the dragonfly algorithm. In addition, the optimal parameters of the support vector machine algorithm were presented using the preceding algorithm to optimize data classification. The performance of the proposed algorithm has been evaluated in comparison with the latest available methods. According to other algorithms, the proposed method is improved by 3.37% and 9.17% accuracy. Accordingly, in terms of accuracy compared to the latest work done, the proposed method has a significant improvement of 34.12%. The proposed method has been tested again on the information received from 7 patients from one of the specialized dialysis clinics in Mashhad.