作者: Linda C. van der Gaag
DOI:
关键词: Joint probability distribution 、 Probability box 、 Probability and statistics 、 Posterior probability 、 Data mining 、 Applied probability 、 Marginal distribution 、 Empirical probability 、 Computer science 、 Mathematical optimization 、 Imprecise probability
摘要: Many AI researchers argue that probability theory is only capable of dealing with uncertainty in situations where a fully specified joint distribution available, and conclude it not suitable for application systems. Probability intervals, however, constitute means expressing incompleteness information. We present method computing interval! probabilities interest from partial specification distribution. Our improves on earlier approaches by all owing independency relation ships between statistical variables to be exploited .