Eliciting hybrid probability-possibility functions and their decision evaluation models

Didier Dubois, Romain Guillaume, Agnès Rico
Proceedings of the Thirteenth International Symposium on Imprecise Probability: Theories and Applications, PMLR 215:200-209, 2023.

Abstract

We focus on a decision tree model under uncertainty using so-called hybrid probability-possibility functions. They allow to handle behaviours lying between possibilistic decision making and probabilistic decision making while keeping the good properties of both approaches namely Dynamic Consistency, Consequentialism and Tree Reduction. We shed light on the various utility functionals in this setting. More precisely, in this paper, we investigate the question of parameterizing the compromise between possibilistic and probabilisic models in different contexts. To this end, we outline elicitation methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v215-dubois23a, title = {Eliciting hybrid probability-possibility functions and their decision evaluation models}, author = {Dubois, Didier and Guillaume, Romain and Rico, Agn\`es}, booktitle = {Proceedings of the Thirteenth International Symposium on Imprecise Probability: Theories and Applications}, pages = {200--209}, year = {2023}, editor = {Miranda, Enrique and Montes, Ignacio and Quaeghebeur, Erik and Vantaggi, Barbara}, volume = {215}, series = {Proceedings of Machine Learning Research}, month = {11--14 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v215/dubois23a/dubois23a.pdf}, url = {https://proceedings.mlr.press/v215/dubois23a.html}, abstract = {We focus on a decision tree model under uncertainty using so-called hybrid probability-possibility functions. They allow to handle behaviours lying between possibilistic decision making and probabilistic decision making while keeping the good properties of both approaches namely Dynamic Consistency, Consequentialism and Tree Reduction. We shed light on the various utility functionals in this setting. More precisely, in this paper, we investigate the question of parameterizing the compromise between possibilistic and probabilisic models in different contexts. To this end, we outline elicitation methods.} }
Endnote
%0 Conference Paper %T Eliciting hybrid probability-possibility functions and their decision evaluation models %A Didier Dubois %A Romain Guillaume %A Agnès Rico %B Proceedings of the Thirteenth International Symposium on Imprecise Probability: Theories and Applications %C Proceedings of Machine Learning Research %D 2023 %E Enrique Miranda %E Ignacio Montes %E Erik Quaeghebeur %E Barbara Vantaggi %F pmlr-v215-dubois23a %I PMLR %P 200--209 %U https://proceedings.mlr.press/v215/dubois23a.html %V 215 %X We focus on a decision tree model under uncertainty using so-called hybrid probability-possibility functions. They allow to handle behaviours lying between possibilistic decision making and probabilistic decision making while keeping the good properties of both approaches namely Dynamic Consistency, Consequentialism and Tree Reduction. We shed light on the various utility functionals in this setting. More precisely, in this paper, we investigate the question of parameterizing the compromise between possibilistic and probabilisic models in different contexts. To this end, we outline elicitation methods.
APA
Dubois, D., Guillaume, R. & Rico, A.. (2023). Eliciting hybrid probability-possibility functions and their decision evaluation models. Proceedings of the Thirteenth International Symposium on Imprecise Probability: Theories and Applications, in Proceedings of Machine Learning Research 215:200-209 Available from https://proceedings.mlr.press/v215/dubois23a.html.

Related Material