Prediction for chronic kidney disease by categorical and non_categorical attributes using different machine learning...

Area of Science:

• Nephrology
• Medical Informatics
• Machine Learning in Healthcare

Background:

• Chronic kidney disease (CKD) presents diagnostic challenges due to asymptomatic progression.
• Early identification of kidney failure is essential for initiating dialysis or transplantation.
• Machine learning models are increasingly utilized for disease prediction and management in healthcare.

Purpose of the Study:

• To develop an accurate early detection model for chronic kidney disease (CKD).
• To evaluate the efficacy of baseline classifiers on different data types (categorical, non-categorical, and combined).
• To enhance classification performance through a majority voting ensemble method.

Main Methods:

• Application of baseline classifiers on categorical and non-categorical attributes separately.
• Integration of classifiers using a majority voting mechanism to combine predictions.
• Comparative analysis against existing models to assess accuracy improvements.

Main Results:

• The proposed model, utilizing baseline classifiers and majority voting, demonstrated a 3% increase in accuracy.
• The ensemble approach showed improved performance in classifying chronic kidney disease.
• The findings support the enhanced accuracy of the developed model for CKD classification.

Conclusions:

• The developed machine learning model offers a promising approach for the early and accurate detection of CKD.
• The majority voting method effectively enhances classification accuracy, aiding clinical decision-making.
• This study contributes to improving diagnostic capabilities for chronic kidney disease, benefiting patient outcomes.