Canonical representation of preferences over a set of probability distributions.
Proceedings of the International Scientific School
"Modelling and Analysis of Safety and Risk in Complex
Systems", St.-Petersburg, June 2004, 248-252.
Building a model of individual preferences is a key for rational
decision-making under uncertainty. Since preference can hardly
be studied completely, its approximation from partially known
preference is very important. The present paper provides a framework
for building such approximation for regular preferences on abstract
partially ordered set, and then applies the results to preferences
on the set of distributions, thus establishing a link to decision-making.