Sensitivity Analysis and Validation Methodologies of Spatially Explicit Agent-Based Modeling

Type: Paper
Theme:
Sponsor Groups: Geographic Information Science and Systems Specialty Group, Spatial Analysis and Modeling Specialty Group
Poster #:
Day: 4/13/2018
Start / End Time: 8:00 AM / 9:40 AM
Room: Gallier A, Sheraton, 4th Floor
Organizers: Jeon-Young Kang, Jared Aldstadt, Arika Ligmann-Zielinska
Chairs: Jeon-Young Kang

Description

Spatially Explicit Agent-based models (ABMs) have become an attractive and efficient method for modelling complex systems composed of autonomous interacting agents. Models are, by necessity, simplifications of the real world. Not only is it unfeasible to include every component of most processes under study, but there is a loss of generality that comes with building a highly-detailed model. The tradeoff of simplification and abstraction may be imprecision or inaccuracy. To account for uncertainties in ABMs introduced by such simplification and abstraction, sensitivity analysis (SA) of model outputs is now recognized as an important part of the modeling process. Model evaluation also involves the conceptual or empirical testing for validating ABMs. Pattern-oriented modelling (POM) is one of the modelling frameworks for evaluating models that focus on more ‘experimental’ simulations (Grimm et al., 2005). It is proposed to use spatial or temporal patterns in the system represented to provide the design, testing, and evaluation of simulation models.
This session aims to address the following questions: How can we evaluate a model’s representation of the real-world system? To what extent can we rely on model predictions? How can we assess if a model replicates the observed pattern?
We are seeking submission on conceptual, methodological and empirical explorations of ABMs. We welcome any topics pertaining to issues of the SA and the validation of spatially explicit ABMs. The proposed topics include (but are not limited to):
• Conceptual methods of sensitivity analysis and ABM validation
• Experimental methods of sensitivity analysis and ABM validation
• Empirical visualization techniques and statistical tests
• Pattern-oriented modelling
• Applications of sensitivity analyses
• Implications of the simulated results from ABMs


Agenda

Type Details Minutes Start Time
Presenter Elisabeth Titis*, Warwick Institute for the Science of Cities, University of Warwick, Rob Procter, Warwick Institute for the Science of Cities at the University of Warwick, Department of Computer Science at the University of Warwick, The Alan Turing Institute in London, Stephen A. Jarvis, Warwick Institute for the Science of Cities at the University of Warwick, Department of Computer Science at the University of Warwick, The Alan Turing Institute in London, Henry Crosby, Warwick Institute for the Science of Cities, University of Warwick, Estimating Interdependences Across Space, Time and Spatial Disparities in Diet and Health in Agent-Based Model for Policy Interventions: The Case of Food Deserts in Greater London/UK 20 8:00 AM
Presenter Jared Aldstadt*, University at Buffalo, Jeon-Young Kang, University at Buffalo, Examining dynamic associations in water, sanitation, and hygiene (WASH) interventions using agent-based models 20 8:20 AM
Presenter Feixue Li*, Department of Geography, Nanjing University, China, Keith C. Clarke, Department of Geography, University of California, Santa Barbara, Santa Barbara, Zhongkai Xie, Department of Geographic Information Science, Nanjing University, China, Manchun Li, Department of Geographic Information Science, Nanjing University, China, Zhenjie Chen, Department of Geographic Information Science, Nanjing University, China, Jian Liang, State Oceanic Administration of China, An agent-based procedure with an embedded agent learning model for urban growth simulation 20 8:40 AM
Presenter Yu Han*, University of Florida, An agent based modeling approach to explore responses of coastal residents on flooding risk 20 9:00 AM
Presenter Arika Ligmann-Zielinska*, Michigan State University, 'Can you fix it?' Using Variance-Based Sensitivity Analysis to Reduce the Input Space of an Agent-Based Model of Land Use Change 20 9:20 AM

To access contact information login