Robustness metrics: How are they calculated and when should they be used?
presentationposted on 2017-12-15, 04:01 authored by Cameron McPhail, Holger Maier, Jan Kwakkel, Matteo Giuliani, Andrea Castelletti, Seth WestraSeth Westra
Environmental modelers have long considered uncertainty, but this has mostly been focused on quantifiable uncertainty around a single projected future. There is now an understanding that multiple plausible futures must be considered where the relative likelihood of each of these projected futures is unknown. Robustness is a way of measuring the performance across multiple plausible futures, however there are many robustness metrics that reflect different aspects of what makes a decision alternative robust.
This research was presented at MODSIM 2017 conference in Hobart, Australia.