Classification of non-chronic and chronic kidney disease using SVM neural networks

Ravindra, B V and Sriraam, N and Geetha, M (2018) Classification of non-chronic and chronic kidney disease using SVM neural networks. International Journal of Engineering & Technology, 7 (1.3). pp. 191-194. ISSN 2394-627X

[img] PDF
4194.pdf - Published Version
Restricted to Registered users only

Download (274kB) | Request a copy


Chronic kidney disease (CKD) refers to the failure of the renal functionalities that leads to the deposition of wastes, electrolytes and other fluids in the body. It is very important to recognize the symptoms that cause the CKD and pathological blood and urine test indicates the key attributes. It is well fact that one has to undergo dialysis due to renal failure. The severity level of disease can be predicted as well as classified using appropriate computer aided quantitative tools. This specific study discusses the classification of chronic and non-chronic kidney disease NCKD using support vector machine (SVM) neural networks. The simulation study makes use of UCI repository CKD datasets with n=400. In order to train to train the attributes of kidney dialysis four cases were considered by including the nominal and numerical values. A radical basis kernel function was employed to train SVM. The performance of the proposed scheme is evaluated in terms of the sensitivity, specificity and classification accuracy. Results reveal an overall classification accuracy of 94.44% was obtained by combining 6 attributes. It can be concluded that the SVM based approach found to be a potential candidate for classification of CKD and NCKD.

Item Type: Article
Subjects: Engineering > MIT Manipal > Computer Science and Engineering
Information Sciences > MCIS Manipal
Depositing User: MIT Library
Date Deposited: 23 Jun 2018 10:42
Last Modified: 23 Jun 2018 10:42

Actions (login required)

View Item View Item