Hello all, I would like to draw attention of those who are interested in decision making with fuzzy possibility and DS belief functions to our works on this topic. The most recent one appeared in this year UAI (see BibTeX reference bellow). With your permission, I'll make a brief sale pitch for this work. Partial consonant belief (pcb) functions class - a subclass of DS belief functions - includes standard probability and fuzzy possibility as extreme cases. An axiomatic decision theory is developed for pcb. The axiomatics is similar to von Neumann-Morgenstern linear utility theory. (In case of standard probability, this axiomatics collapses to vNM axiomatics.) In addition to risk attitude notion examined in vNM theory, this theory also handles ambiguity (uncertainty) attitude. In the paper cited there is a simple example showing how Ellsberg's paradox is resolved in the framework of our theory.
I'm interested in any comments and critiques that you have. Thanks, Phan Giang, Ph.D., Staff Scientist Siemens Medical Solutions Inc. Computer-Aided Diagnosis Solutions Group Mail stop H32 51 Valley Stream Pkwy Malvern PA 19355 Phone: (610) 448-4324 Email: [EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]> @INPROCEEDINGS{GS03, AUTHOR="Giang, P. H. and Shenoy, P. P.", EDITOR="Meek, Chris and Kjaerulff, Uffe", TITLE ="Decision Making with Partially Consonant Belief Functions", BOOKTITLE = "Uncertainty in Artificial Intelligence: Proceedings of the Nineteenth Conference (UAI--2003)", PUBLISHER = "Morgan Kaufmann", ADDRESS = "San Francisco, CA", YEAR = "2003", PAGES = "", MONTH="08", }
