Research: Generic Steps for Developing Simulation Models

ITS 832 Chapter 6

Features and Added Value of Simulation Models Using Different Modelling

Approaches Supporting Policy-Making Information Technology in a Global Economy

Professor Michael Solomon

Introduction

• Simulation Models in policy-making – foundations • eGovPoliNet

• International multidisciplinary policy community in ICT

• Selected Modeling approaches • VirSim – Pandemic policy

• microSim – Swedish population

• MEL-C – Early Life-course

• Ocopomo’s Kosice Case – Energy policy

• SKIN – Dynamic systems component interaction

Foundations of Simulation modeling

• Simulation model • Smaller, less detailed, less complex (or all)

• Computer software • Approximates real-world behavior

• Benefits • Easier, simpler than monitoring reality • Possibly the only feasible way to “play out” a scenario

• Approaches discussed • System dynamics • Agent-based modeling (ABM) • Micro-simulation

Steps in Developing Simulation Models

Simulation Models Examined

VirSim

• A Model to Support Pandemic Policy-Making • Simulates the spread of pandemic influenza

• Goal • Determine the optimal time and duration of school closings to affect

influenza spread

• System dynamics model • Separates population into 3 segments

• Younger than 20 years old • 20 – 59 years old • 60 years old and older

• No environmental features considered • Only input data for Sweden

MicroSim

• Micro-simulation Model • Modeling the Swedish Population

• Goal • Determine how multiple behavior features affect influenza

spread

• Micro-simulation model • More granular than VirSim • Focused only on Sweden • Robust for intended population

MEL-C

• Modeling the Early Life-Course • Knowledge-based inquiry tool With Intervention modeling (KIWI)

• Goal • Identify social development milestones in early life that most affect

later outcomes

• Health, nutrition, education, living conditions, etc. • Micro-simulation model • Generic applicability • Limited by range of options

• Evidence-based • Not very flexible when considering untested approaches

Ocopomo’s Kosice Case

• Kosice self-governing region energy policy simulation • Goal

• Develop better energy policy • And measure policy effectiveness

• House insulation and renewable energy sources

• ABM model • Model is geographically anchored

• Difficult to apply to other regions • Many geographic features

• Stakeholder engagement is key

SKIN

• Simulating Knowledge Dynamics in Innovation Networks • Goal

• Improve innovation through interactions • ABM model • Based on general market model • Agents are both

• Sellers (providers) • Buyers (consumers)

• Agents consider dynamic interaction • Modify behavior to improve innovation • i.e. sell more or buy better

Summary

• Examined five models built on three approaches • VirSim – System dynamics

• MicroSim - Microsimulation

• MEL-C - Microsimulation

• Ocopomo’s Kosice Case - ABM

• SKIN – ABM

• Each approach has advantages and limitations • Simulations allow multiple models to be investigated

• Without real-world consequences