Pemodelan Epidemiologi Berbasis Data untuk Memprediksi Penyebaran Penyakit Menular di Wilayah Perkotaan

Abstract

Infectious diseases remain a major challenge in public health, particularly in urban areas characterized by high population density, dynamic mobility, and intensive social interactions. These conditions increase the risk of rapid and widespread disease transmission, requiring effective prevention and control strategies. This study aims to develop a data-driven epidemiological model to predict the spread patterns of infectious diseases in urban settings. The approach integrates data analysis techniques with the Susceptible-Infected-Recovered (SIR) model to represent disease transmission dynamics based on population characteristics and environmental factors. The dataset includes daily case counts, population density, mobility levels, vaccination rates, and demographic information obtained from health institutions and public data sources. The research process involves data collection, data cleaning, statistical exploration, model development, and validation using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). Simulation of various public health intervention scenarios is also conducted to assess policy impacts on transmission rates. The results show that the model provides accurate short- and medium-term predictions. Population density and mobility significantly influence case increases, while vaccination and mobility restrictions effectively reduce active cases. This model supports evidence-based decision-making and early warning systems in urban public health management.