Predicting Disease Outbreaks with Spatio-Temporal Machine Learning Models and Epidemiological Data

Accurate prediction of infectious disease outbreaks are vital for effective public health planning and intervention. This paper proposes a novel spatio-temporal prediction framework that integrates Graph Convolutional Networks (GCNs) with Long Short-Term Memory (LSTM) networks to model the complex spatial and temporal dynamics of disease spread. By leveraging multi-source data—including epidemiological time-series, climatic variables, demographic distributions, and human mobility patterns—the model learns region-wise interdependencies and sequential trends in disease progression. The spatial component, implemented via GCN, captures inter-regional transmission patterns, while the temporal component, powered by LSTM, models evolving outbreak trends within each region. Experimental evaluation on real-world datasets demonstrates that the proposed GCN+LSTM model significantly outperforms traditional models such as ARIMA, Random Forest, and standalone LSTM in terms of MAE, RMSE, and F1-score. It also meets high classification performance at low false negative rate and therefore very suitable for real-time outbreak surveillance and early warning system.