Author

Eric Lavin

Publication Date

2018

Document Type

Dissertation/Thesis

First Advisor

Giabbanelli, Philippe J.

Degree Name

M.S. (Master of Science)

Legacy Department

Department of Computer Science

LCSH

Computer science

Abstract

To study complex phenomena, modelers have used a wide variety of modeling techniques (e.g. agent-based modeling). While these are powerful and useful modeling methods, they are not perfect. To improve our models of complex problems, we need to apply methods that handle uncertainty such as Fuzzy Cognitive Maps (FCM). This thesis will show that while we use simulation models to evaluate complex phenomena, we do not apply FCMs in these evaluations very often. While they are commonly used in participatory modeling to handle uncertainty in small models, this thesis proposes and evaluates a new method to handle uncertainty in much larger models using FCMs. Specifically, we use parallel Design of Experiments (DoE) and demonstrate that previously published large models can be simplified. We will present a design of experiments which will extend the ability of Fuzzy Grey Cognitive Maps (FGCM) to better handle uncertainty, and identify the main factors that determine the simulation output. We also explore an approximation to allow for execution within a time limit. Beyond handling uncertainty, this thesis will examine whether we need to run several simulations to compare different FCMs, or if we can just compare their structures as graphs. This work is limited by the small number of FGCMs currently published. In the future, our work can be used to create a hybrid model for agent-based models with FCMs, or to identify the rules necessary to create a heterogeneous population of agents.

Comments

Advisors: Philippe J. Giabbanelli.||Committee members: Hamed Alhoori; Minmei Hou; Scott Rosen.||Includes illustrations.||Includes bibliographical references.

Extent

155 pages

Language

eng

Publisher

Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type

Text

Share

COinS