HYPERPARAMETER TUNING METHOD OF EXTREME LEARNING MACHINE (ELM) USING GRIDSEARCHCV IN CLASSIFICATION OF PNEUMONIA IN TODDLERS

Abstract

Pneumonia is a disease that is susceptible to attack toddlers. According to data from the Ministry of Health, the cause of under-five mortality due to pneumonia is number 2 of all under-five deaths. The dataset used is pneumonia disease data at the MTBS Health Center of East Martapura Health Center. The classification method in this study uses the Extreme Learning Machine (ELM) method. The classification process starts from SMOTE upsampling to balance the class, then parameter tunning is performed using GridsearchCV on the hidden layer neurons, then classification is carried out using the ELM method using the Triangular Basis activation function by comparing the test datasets 90:10, 80:20, 70:30, 60:40 and 50:50. This study provides the best performance results with an accuracy of 86.36%, the ratio of training and test data is 90:10 and 3 neurons hidden layer.