Fig. 2: Area under ROC (left) and PR (right) curves from forecasts of conflict incidence in Africa at different administrative levels: individual predictive power of indicator groups.
From: Extreme weather impacts do not improve conflict predictions in Africa
These figures show the individual model performances—ROC and PR curve results—of conflict incidence predictions made using a Generalized Random Forest (GRF) of four indicator groups: conflict history (conf-green), governance indicators (gov-gold), extreme weather events (ewi-red), and socioeconomic indicators (socecon-blue). The predictions are conducted at different administrative levels from the most aggregate (level 0 is the national level) to the most granular (level 2). Battles from the Uppsala Conflict Data Program (UCDP) are used as an outcome variable.
