Predictive Analysis of Malaria Cases in Indonesia Using Machine Learning

Abstract

Malaria continues to pose a significant public health challenge globally, with Indonesia being among the countries most affected by the disease. Despite extensive efforts to control malaria transmission, the disease remains endemic in various regions, leading to substantial morbidity and mortality. Accurate prediction of malaria cases is crucial for guiding effective prevention and control strategies, particularly in resource-limited settings. This study investigates the application of machine learning (ML) techniques to predict malaria incidence in Indonesia, leveraging climatic, epidemiological, and socioeconomic data. Three ML algorithms, namely Random Forest, Support Vector Machine (SVM), and Artificial Neural Networks (ANN), are employed and evaluated for their predictive capabilities. The study spans from 2010 to 2021, incorporating diverse datasets from the Indonesian Meteorological, Climatological, and Geophysical Agency (BMKG), the Ministry of Health of Indonesia, and the Indonesian Bureau of Statistics (BPS). Results indicate that the ML models exhibit strong predictive performance, with Random Forest demonstrating the highest accuracy. The integration of multidimensional data sources enhances the robustness of the predictive models, enabling the identification of spatiotemporal patterns in malaria transmission dynamics. The findings underscore the potential of ML-based approaches in improving malaria surveillance and control efforts in Indonesia, offering valuable insights for public health decision-makers and stakeholders. Moreover, the study highlights the importance of data quality, model refinement, and interdisciplinary collaboration in addressing complex public health challenges such as malaria. By harnessing the power of advanced analytics and innovative methodologies, this research contributes to the ongoing efforts to combat malaria and alleviate its burden on communities and healthcare systems in Indonesia and beyond.