Forecasting Political and Economic Crisis: Social Science Meets Machine Learning

Political and economic crisis (like armed conflict, famines, expropriations or banking crisis) are socially costly and economically damaging. Organisations both in the public sector and the private sector therefore have an interest in forecasting these events. Crisis forecasting is also an old problem in Social Science and methods in academia vary from expert opinion to fully automated systems. With the development of the internet and social media and recent advances in machine learning some of the old problems have been approached in new ways.

Academic research on crisis forecasting is important partly to complement and support forecasting efforts within public and private organizations, but also to understand and challenge such forecasting since they are often done in non-transparent ways and may have an important impact on the crisis events themselves.

This workshop invites cutting edge research that tries to forecast or discusses forecasting of rare negative events using quantitative methods of evaluating risk. We explicitly invite researchers from all fields, inside and outside of academia and using methods spanning from expert opinion to machine learning. The goal of the workshop is to foster an exchange about methods across fields and between operational systems and academic models.

Workshop organizers

• Håvard Hegre (Uppsala University)
• Laura Mayoral (IAE-CSIC and BSE)
• Hannes Mueller (IAE-CSIC and BSE)
• Omiros Papaspiliopoulos (ICREA-UPF and BSE)