Prediction of Chronic Kidney Disease using SVM and CNN

Abstract

Chronic kidney disease is one of the deadliest diseases today and it is vital to have a good diagnosis as soon as possible. In medical treatment, machine learning has been reported to be effective. A doctor can diagnose the disease early by using machine learning classifier algorithms. This study investigated the chronic disease prognosis of this concept. Disease data was taken from the University of California, Irvine. Other measurement algorithms used in this study include C5.0, Chi-square automatic interaction detector, line extraction, SVM line with L1 and L2 flap, and neural network random tree. The database was also submitted to a feature selection program that merited the database. Scores are computer generated for each category segment using the following methods: Full Version, (ii) Link-Based Feature Selection, (iii) Folder Feature Selection, (iv) Minimal Collapse and Selected Optional Retrospective Features, (v) integrated small oversampling method with very small reduction features and selected bias on the selected operator, and (vi) how to do multiple samples combined with full functions. In the full multi-sample processing process, the findings show that L2-loaded LSVM has a very high accuracy of 96.86 percent. The graph shows the results of different methods, as well as precision, precision, recall, F-score, area under the curve, and GINI coefficient. The minimum absolute reduction and selection regression operation selected features using the synthetic minority oversampling approach produced the best results after using the synthetic minority oversampling method with full features. The support vector machine achieved a high accuracy of 96.46 percent in the process of making very large samples with very small turndowns and selected operator features. Machine learning methods used with convolutional neural networks and SVM classifier models on the same database, with 96.7 percent of high-definition support machine models and networks are used.