CLASSIFICATION OF K-NEAREST NEIGHBOR (K-NN) AND CONVOLUTIONAL NEURAL NETWORK (CNN) FOR THE IDENTIFICATION OF BRONCHITIS DISEASE IN TODDLERS USING GLCM FEATURE EXTRACTION BASED ON THORAX X-RAY IMAGES

Abstract

K-Nearest Neighbor (K-NN) is a classification method that seeks the majority class from the k-nearest neighbors of a sample to be classified. Meanwhile, Convolutional Neural Network (CNN) is a type of artificial neural network specifically designed to recognize patterns in image data. The features are then extracted using GLCM (Gray Level Co-occurrence Matrix) from Thorax X-Ray images. This research aims to develop two classification approaches, namely K-Nearest Neighbor (K-NN) and Convolutional Neural Network (CNN), to detect bronchitis disease in toddlers based on Thorax X-Ray images. Feature extraction based on the Gray Level Co-occurrence Matrix (GLCM) is used to transform images into numerical features that can be processed by classification algorithms. The results from both methods will be combined based on various evaluation metrics, such as accuracy, precision, recall, F1-score, etc