Chronic-Kidney-Disease-Detection

Introduction:

This project aims to classify whether a patient has chronic kidney disease or not. We will develop three neural networks:

• Perceptron Neural Network
• Back propagation Neural Network
• Momentum Neural Network

We will use kidney disease dataset from Kaggle. It has 200 rows with 20 features like red blood cells, pedal edema, sugar, etc.

Results:

• The perceptron algorithm will not be able to correctly classify all examples, with 200 time of iteration and Learning Rate 0.01 our perceptron got a 82.32% test accuracy.
• The Back Propagation network with 500 time of iteration, Learning Rate 0.3 and 3 Hidden Layers our algorithm got a 99.50% test accuracy.
• The momentum algorithm with 50 time of iteration, Learning Rate 0.05 and 5 Hidden Layers our algorithm got a 95.83% test accuracy.