CrossLag: Predicting Major Dengue Outbreaks with a Domain Knowledge Informed Transformer
This work addresses the problem of timely public warnings for major dengue outbreaks, which is critical for public health, but it appears incremental as it builds on existing transformer methods.
The paper tackles the challenge of predicting major dengue outbreaks by introducing CrossLag, an environmentally informed attention mechanism that incorporates lagging signals into a transformer architecture, resulting in outperforming the baseline TimeXer model in detecting and predicting outbreaks over a 24-week window.
A variety of models have been developed to forecast dengue cases to date. However, it remains a challenge to predict major dengue outbreaks that need timely public warnings the most. In this paper, we introduce CrossLag, an environmentally informed attention that allows for the incorporation of lagging endogenous signals behind the significant events in the exogenous data into the architecture of the transformer at low parameter counts. Outbreaks typically lag behind major changes in climate and oceanic anomalies. We use TimeXer, a recent general-purpose transformer distinguishing exogenous-endogenous inputs, as the baseline for this study. Our proposed model outperforms TimeXer by a considerable margin in detecting and predicting major outbreaks in Singapore dengue data over a 24-week prediction window.