作者: George Wright
DOI: 10.1007/978-94-017-1590-4_12
关键词: Probability assessment 、 Probabilistic logic 、 Probabilistic forecasting 、 Selection (genetic algorithm) 、 Machine learning 、 Calibration (statistics) 、 Computer science 、 Artificial intelligence 、 General knowledge
摘要: This paper introduces the concept of calibration which refers to external correspondence between assessed subjective probabilities and hit-rate. Evidence is presented that psychological processes involved in forecasting future events are different from those evaluating truth one’s own knowledge. The implication results research has investigated probability assessments given general knowledge questions will not generalize real world where uncertainty located future. Finally, directions for on probabilistic sketched out. should lead effective evaluation forecasts selection forecasters.