A Novel Prediction Analytics Science of Kidney Disease based on Neural Network

Abstract

Chronic kidney disease (CKD) is a condition wherein people face a progressive deterioration of kidney function over a period. In medical fields, the prediction of this illness is perhaps one of the main critical things. Various automated ML algorithms can aid in the prediction of the patient's kidney condition and treatment. The importance of this disease's early determination approach requires attention, mainly in developing nations, where it is often identified late. So, the most sensible solution is to find a feasible solution and decreasing the disadvantages. The proposed system aims to introduce a novel method to use Feature selection by using Pearson correlation and Univariate selection. The implications of using clinical characteristics to detect victims with CKD using support vector machines procedure are investigated in this survey. For the dataset obtained out of a generic database, separate metrics had been used to compare algorithms. The results will then be compared with Neural networks. Through the proposed model, accurate prediction for CKD will be done, along with specificity and sensitivity. Proper comparison will also be done between existing and proposed algorithms.