Modeling Precursors for Event Forecasting via Nested Multi-Instance Learning

Forecasting large-scale societal events like civil unrest movements, disease outbreaks, and elections is an important and challenging problem. From the perspective of human analysts and policy makers,