par Smets, Philippe
Référence Colloquium - Institution of Electrical Engineers, 86
Publication Publié, 1990
Article révisé par les pairs
Résumé : The transferable belief model (TBM) is a model to quantify personal subjective belief. It covers the same domain as Bayesian model except it is based on belief functions instead of probability functions. The TBM is perfectly suited to describe any level of partial belief up to the total ignorance. An example concerning breakable sensors for which one knows only its behaviour when it is in working condition but ignores completely how it behaves when broken is described.