Comparative Study on Chronic Kidney Disease Detection Using Tree-Based Models

EasyChair Preprint 7854

Abstract

Chronic Kidney disease is the slow, progressive deterioration of the functioning of the kidneys. This impairment is irreversible if it reaches the later stages and hence demands early detection and treatment to ensure prolonged functioning of the kidney.

In this project, we have developed  ML model to classify whether a person has Chronic Kidney Disease (CKD) or not. The idea of this study is to calculate the performance of various decision tree-based learning algorithms and to compare their accuracies. In our work, we have used the dataset from UCI which contains real-time data. The UCI’s CKD dataset has 400 entries and has missing/noise information. It has 250 patients that have CKD and 150 that have non-CKD, consisting of attributes like age, blood pressure, specific gravity, etc. A total of 5 Tree Bases ML classifiers have been used achieving accuracies as high as 99%.

Keyphrases: Chronic Kidney Disease, Machine Learning Algorithms, Tree-based algorithms, ensemble methods

@booklet{EasyChair:7854,
  author    = {Siddhartha Krishna and Maanvi Keesara and Mary Subaja},
  title     = {Comparative Study on Chronic Kidney Disease Detection Using Tree-Based Models},
  howpublished = {EasyChair Preprint 7854},
  year      = {EasyChair, 2022}}