CHRONIC KIDNEY DISEASE DIAGNOSIS SYSTEM USING MACHINE LEARNING

Abstract

Chronic Kidney Disease (CKD) is a progressive
and life-threatening condition that affects a
significant portion of the global population and
often remains undetected in its early stages due to
the absence of noticeable symptoms. Early
diagnosis plays a crucial role in preventing severe
complications such as kidney failure,
cardiovascular diseases, and metabolic disorders. In
recent years, machine learning (ML) techniques
have emerged as powerful tools for medical
diagnosis due to their ability to analyze complex
datasets and identify hidden patterns. This project
presents a machine learning-based CKD diagnosis
system that utilizes clinical data obtained from the
UCI Machine Learning Repository. The dataset
contains missing values, which are effectively
handled using K-Nearest Neighbor (KNN)
imputation to ensure data completeness and
reliability. Multiple machine learning algorithms,
including Logistic Regression, Random Forest,
Support Vector Machine, K-Nearest Neighbor,
Naïve Bayes, and Feed Forward Neural Network,
are implemented to classify patients as CKD or
non-CKD. Among these models, Random Forest
achieves the highest accuracy due to its robustness
and ability to handle nonlinear relationships.
Furthermore, a hybrid ensemble model combining
Logistic Regression and Random Forest using a
perceptron approach is proposed to enhance
predictive performance. The system demonstrates
high accuracy and reliability, making it suitable for
real-world clinical applications. This approach
enables early detection, improves decision-making,
and reduces healthcare costs by assisting medical
professionals in diagnosing CKD effectively.