Lower Previsions for Unbounded Random Variables
Matthias C. M. Troffaes, Gert de Cooman · Advances in intelligent and soft computing · 2002
In order to generalise Walley’s theory of lower previsions, which are real-valued maps on bounded random variables, to arbitrary random variables, we introduce extended lower previsions as extended real-valued maps on arbitrary, not necessarily bounded, random variables. We suggest and motivate conditions for avoiding sure loss, coherence and linearity, we construct a natural extension, and we suggest a way to generalise some of the more advanced topological results from the existing theory of lower previsions. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.