Optimization of C4.5 Algorithm Performance Using Particle Swarm Optimization in Predicting Stunting Risk

Abstract

Stunting is a serious global health problem, especially in developing countries. It is caused by chronic malnutrition in children, especially toddlers, which inhibits physical and cognitive growth. Stunting also has the potential to reduce quality of life and productivity in the future. Therefore, early detection of stunting risk is crucial so that appropriate interventions can be provided. Currently, data mining-based classification methods, such as the C4.5 algorithm, have been widely used to predict stunting risk. However, the performance of the C4.5 algorithm in terms of accuracy and efficiency is still lacking, especially in attribute selection and parameter settings. This research aims to improve the accuracy of the C4.5 algorithm in predicting stunting risk by implementing Particle Swarm Optimization (PSO) as an optimization technique. PSO is chosen because of its ability to find optimal solutions quickly and efficiently through the principles of particle social behavior. By using PSO, this research is expected to optimize the attribute selection process and parameter settings in the C4.5 algorithm, so as to produce a more accurate classification model in detecting stunting risk. The result of this research is a significant increase in prediction accuracy compared to the use of the C4.5 algorithm without optimization, so that the resulting model can be a more reliable tool for governments, health institutions, and other policy makers in designing interventions and strategies to overcome stunting.