การเปรียบเทียบประสิทธิภาพการจำแนกประเภทข้อมูลความเสี่ยงการเป็นโรคไต ด้วยเทคนิคการทำเหมืองข้อมูล

Abstract

This research aims to apply classification data mining techniques to predict the risk of kidney disease. and compare the performance to achieve the most efficient algorithm with three different data mining algorithms: neural network, decision tree, and Bayesian learning. The dataset used in the study was the early chronic kidney disease dataset. that are archived in a UCI machine learning database repository. containing 400 datasets and 24 factors. And developing a classification model in Python. The results showed that the best method for kidney disease classification was the decision tree method that achieved an accuracy of 99.17%, precision 99.00%, recall 99.00% and 99.00% F-measure.