Unimodal density estimation with applications in expert elicitation and decision making under conditions of uncertainty
by Yacov Salamon
Abstract: A wide range of applied problems may be reduced to an under-determined inverse problem of the form: find the "best" density \(f\) subject to a set of constraints. The problem of choosing the best distribution to represent the data elicited from an expert is one such example. In my completion seminar I will discuss our approach to solving such problems when the set of constraints includes unimodality. I will then illustrate our results in the context of expert elicitation. Finally, I will outline a method for incorporating expert elicited data in decision making under conditions of uncertainty. The method is applied in the context of environmental management using sensitivity analysis of matrix population models.