Publication

Abstract : Chronic kidney disease, a long-term kidney disorder that leads to high cardiovascular disease, is mainly caused by diabetes, high blood pressure, obesity, and urinary infections. Kidneys are vital organs in the human body, so early detection of chronic kidney disorders is very necessary in order to startthe clinical process. This research mainly aims at categorizing whether the patient diagnosed has chronic kidney disease from the 24 attributes and comparing them with the values in the dataset to start the medical treatments at the earliest. This study has developed and implemented multiple models and the binary classification neural network got the most accurate result with an accuracy of 99.23%. We also noticed a very low loss value, which is 0.023%, due to the ReLu activation layer function in the hidden layer and the hard sigmoid and sigmoid functions in the output layers. Along with this, the other models deployed include Decision Tree, Extra Tree and Support Vector Machine models in order to get the most accurate model for predicting chronic kidney disease from the provided dataset. The accuracy records of the employed models are Extra Tree (98.75%), Decision Tree (95.0%), and Support Vector Machine (93.75%).