Comparative Study of Chronic Kidney Disease Classification by Using Logistic Regression and Decision Tree



Lim, Wei Yang (2022) Comparative Study of Chronic Kidney Disease Classification by Using Logistic Regression and Decision Tree. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

[img] Text
Lim Wei Yang_FullText.pdf
Restricted to Registered users only

Download (4MB)


Chronic Kidney Disease (CKD), as known as chronic renal disease. Chronic kidney disease would bring you a bunch of serious symptoms that may hurt you or death. Doctors develop a complicated experiment to classify the patient with CKD or without CKD. Early detection and treatment can usually prevent the deterioration of chronic kidney disease. Data mining is the process used for knowledge exploration from the endless databases. The job of data mining is to make use of past data, to discover the original pattern and make better decisions. Machine learning has already established practical application in the real world such as analyzing medical science outcomes, social media, financial problems and so on. Various data mining classification approaches and machine learning algorithms are applied on the prediction of chronic disease. The objective of this case study is to find the suitable model and the key factor for CKD to help doctors reduce their working time on predicting CKD. The aim of this work is to make a model performance comparison between 3 supervised machine learning such as logistic egression, decision tree and XGBoost classifier based on accuracy, precision, recall, f1-score, AUC score and Training time on CKD prediction. From these experimental results it is observed that XGBoost classifier is the best machine learning to help doctor to predict CKD with average accuracy score of 98.3611%, average precision score of 99.8011%, average recall score of 97.22%, average F1 score or 98.4822% and average AUC score of 98.4989%. Not only that, the key factor that I found in experimental results is hemoglobin with Feature Score of 60 and specific gravity feature Score of 55.

Item Type: Final Year Project
Subjects: Science > Mathematics
Medicine > Internal medicine
Faculties: Faculty of Computing and Information Technology > Bachelor of Science (Honours) in Management Mathematics with Computing
Depositing User: Library Staff
Date Deposited: 17 Aug 2022 03:19
Last Modified: 17 Aug 2022 03:19