search

JannisHoch / copro

(ML) model for computing conflict risk from climate, environmental, and societal drivers.
https://copro.readthedocs.io/en/latest/
MIT License
6 stars 0 forks source link

assess model quality re. conf #80

Closed JannisHoch closed 4 years ago

JannisHoch commented 4 years ago

we know after n model repetitions how many predicitons were made per polygon, how often the model prediction was correct, for both conflict and non-conflict as well as for conflict only.

this gives an overall good impression of model performance.

however, how do we show which polygons are now predicted to be 'at risk'? because with multiple predictions made per polygon, no every prediction is only 'conflict' or only 'non-conflict', so we will have to deal with a melange of predictions. what are good means to visualize this per polygon?

JannisHoch commented 4 years ago

solved in https://github.com/JannisHoch/conflict_model/commit/6f780ee1e7e4de673ba8966410e65b895c13c589 by plotting the chance of conflict. together with averge prediction accuracy this should give a good picture of which polygons are predicted to be at risk and with which certainty the model can say that.