ReBoost: A Robust Chronic Kidney Disease detection using Probability Reweighted AdaBoost

Abstract

Kidney act as one of the important organ of the body. Kidney diseases are needed to be
treated in the early stages to protect the Human life.Chronic kidney diseases slowly impact
the functionality of kidneys that make severe damage to kidneys if untreated. Chronic kidney
diseases affect other organs of the body and create life threatening problems. Early
prediction of kidney diseases can save the life of the person and reduces the financial cost
takes for treatment purposes. The proposed approach is focused on developing a robust
algorithm named probability watered Ada boost to predict the chronic kidney disease based
on various record and information. The adaptive weight calculation of ADO boost algorithm
is modified here to provide the reduction of incorrectly classified values on Kidney Disease
detection. The main purpose of the system is to reduce the first prediction rate. Non chronic
kidney disease is used to compare performance of various existing algorithms such as Naive
Bayes algorithm decisions etc. The presented result is compared the existing system and
achieve the accuracy of 99%.