<p>Pneumonia is an acute respiratory disease that attacks alveoli and distal bronchioles, caused by bacterial, fungal, or parasitic infections. Its incidence is influenced by both internal factors (e.g., age, smoking status, and comorbidities) and external factors, including temperature, humidity, rainfall, air pollution, and sunshine duration. In densely populated urban areas such as Jakarta, Indonesia, these environmental factors are expected to exacerbate pneumonia cases. This study proposes an integration-based modeling framework for analyzing pneumonia cases by incorporating environmental factors and interregional dependencies. A multilayer perceptron (MLP) is first used to capture the complex nonlinear relationship between environmental variables and pneumonia cases in each region. Then, to account for residual interregional dependencies, a vine copula-based error correction mechanism is integrated to enable flexible modeling of high-dimensional dependency structures using C-vine and D-vine copulas. The main contribution of this study lies in the integration of machine learning and dependency modeling through a vine copula-based error correction approach, which can enhance the model’s ability to capture both nonlinear effects and interregional dependencies. Our proposed model is applied to pneumonia and environmental data from Jakarta. The results show that our proposed model improves the prediction accuracy of the MLP by approximately 0.1–2%. Furthermore, the findings reveal significant interregional dependencies and suggest that environmental factors contribute to the variation in pneumonia cases. These results highlight the importance of incorporating environmental variability and spatial dependencies into epidemiological modeling, which can provide valuable insights for public health policy and region-specific intervention strategies.</p>