Kidney disease diagnosis model using support vector machine

Background and aims: Chronic kidney disease means that the human kidneys are damagedand cannot filter the blood as it should, and this damage can cause the creation of toxins in thebody. This disease affects about ۱۰% of the world’s population and its prevalence is increasing.Nowadays, due to the abundant access to various types of medical data, different data miningand artificial intelligence techniques can be used to analyze these data. These analyzes can helpimprove the quality of medical services. The aim of the current research is to use machine learningtechniques to help diagnose chronic kidney disease and to evaluate the effectiveness of thesetechniques compared to other existing techniques.Method: This study is descriptive-analytical. The data used in this research were extracted from۳۰۰ patients and non-patients in Tabriz hospitals. These data were first pre-processed in the Pythonenvironment and removed from noise and outlying observations. Then, support vector machine,multilayer perceptron and decision tree algorithms were used to classify the data. Accuracy,Recall and Precision evaluation criteria were calculated to evaluate the performance of thesecategories.Results: According to the calculated evaluation criteria, for the support vector machine algorithm,the values of Accuracy, Recall and Precision criteria were obtained as ۰.۹۷۴, ۰.۹۴, and۰.۹۶۲, respectively. The findings indicate the better performance of the support vector machinealgorithm in terms of accuracy.Conclusion: The results obtained from the present study indicate the very favorable efficiencyof machine learning techniques in the diagnosis of chronic kidney disease. The use of these techniquescan facilitate the diagnosis and treatment of these patients and increase the probability ofpeople’s recovery.

کلیدواژه ها:

Kidney disease ، diagnosis ، support vector machine ، artificial intelligence ، machine learning