Premium
Informing the Delineation of Input Uncertainty Space in Exploratory Modelling Using a Heuristic Approach
Author(s) -
Moallemi Enayat A.,
Elsawah Sondoss,
Ryan Michael J.
Publication year - 2018
Publication title -
insight
Language(s) - English
Resource type - Journals
eISSN - 2156-4868
pISSN - 2156-485X
DOI - 10.1002/inst.12209
Subject(s) - heuristic , space (punctuation) , computer science , process (computing) , exploratory research , exploratory analysis , operations research , industrial engineering , mathematical optimization , artificial intelligence , mathematics , data science , engineering , sociology , anthropology , operating system
Exploratory modelling is an emerging approach which can address the challenge of model‐based decision making in dealing with input model uncertainties. Exploratory modelling samples from an input uncertainty space and generates extensive computational experiments to analyse possible model behaviours in an output solution space. The way that the input uncertainty space is delineated influences the results of exploratory modelling and its computational cost. In this article, we show the statistical significance of the implication of the size of an input uncertainty space on the resulted output solution space. We also propose a heuristic approach which informs the delineation of input uncertainties by screening the relevant model behaviour in the solution space. An illustrative example of an aircraft fleet management system is used to demonstrate the implementation of our approach in practice. We conclude that the delineation of input uncertainty space can be a way to control simulations in exploratory modelling and to enhance the efficiency of the exploration process and the confidence of the final results.