In which case can you define what an error interval on a probability estimate means? :)
I think probabilities can either be well calibrated or not. (Which then of course begs the question of how you define being well calibrated for probabilities of events that happen very rarely). An error interval does not seem to make too much sense.
I think probabilities can either be well calibrated or not. (Which then of course begs the question of how you define being well calibrated for probabilities of events that happen very rarely). An error interval does not seem to make too much sense.