AI-Based Prediction of Dengue Incidence using Climatic, Environmental, and Socio-Demographic Factors: An Ensemble Random Forest Approach with Agile System Development

Background Dengue fever remains a major public health concern across tropical regions, particularly in Indonesia, where climatic, environmental, and socio-demographic factors interact in complex and dynamic ways. Conventional forecasting systems often fail to adapt to these changing conditions, underscoring the need for intelligent and flexible predictive frameworks. This study aimed to develop an adaptive AI–Agile framework using an ensemble Random Forest (RF) model to predict monthly dengue incidence by integrating climatic, environmental, and socio-demographic variables in Yogyakarta, Indonesia. Methods An ensemble RF model was constructed within an Agile System Development framework to forecast dengue incidence from 2017 to 2022. Multi-domain datasets encompassing climatic (temperature, rainfall, humidity, wind speed, pressure), environmental (built-up area, vegetation, stagnant water, agricultural land), and socio-demographic (population density) factors were incorporated. A Negative Binomial Generalised Linear Model (GLM) was initially employed to identify significant predictors before applying the non-linear RF model for enhanced accuracy and adaptability.. Results The RF model demonstrated strong predictive performance (R² = 0.86, RMSE = 5.72), outperforming the GLM baseline (R² = 0.64). Rainfall with a one-month lag, relative humidity, and temperature were dominant climatic predictors, while built-up area and population density had substantial effects. The Agile approach enabled iterative refinement and seamless data integration, improving responsiveness to emerging transmission patterns. Conclusion The proposed AI–Agile framework effectively integrates multi-domain predictors, enhancing dengue forecasting accuracy and interpretability. This adaptive system provides a scalable foundation for early warning and data-driven vector control in rapidly changing tropical environments.