Simulation and Agent-Based Modelling

Time/Room

  • Tuesday and Friday 08:30-10:00
  • E 114

First Session

  • 24/10/2006, 08:30

Lecturer

  • Klaus G. Troitzsch

Intended Audience

  • see text (WIKT03=WIKT07)

Assessment

  •  

Related Event

Overview

This will be a module on the techniques of developing simulations to help with the exploration and understanding of social and economic issues. It will provide a rationale for using simulation in economics and the social sciences and outline a number of approaches to social simulation at a level of detail that would enable participants to understand the literature and, for some selected approaches, to develop their own simulations. The workshop covers the basics of modelling and simulation in economics and the social sciences from different points of view (mathematics, computer science, philosophy of science) and of seven different approaches to computer simulation in economics and the social sciences.

The aim of this module is a broad introduction into all approaches to simulation in the social sciences. Being not only a ZUMA workshop, but also part of the computer science and information management programmes of the Computer Science Faculty of Koblenz-Landau University (KLU), this module is designed for students in their 5th or 7th semester (for the first time in 2005, it is also part of the faculty's first summer academy). It covers the basics of modelling and simulation in the social sciences from different points of view (mathematics, computer science, philosophy of science) and of seven different approaches to computer simulation in the social sciences.

By the end of this module, a student should understand

  • what simulation is good for in the social sciences and which steps should be taken to arrive at a useful computer simulation and he or she should know
  • which approaches have been followed by social scientist in the past decades, what the aims of these approaches were and which advantages and shortcomings these approaches have.

    Moreover, students should be able to make use of a number of different simulation tools and have gained some experience in designing their own models.

Syllabus

  1. Simulation and Social Science — history, taxonomy, motives, simulation from a philosophy of science point of view (slides in German): 24-27/10/2006
  2. Simulation as a Method — logic of simulation, stages of simulation-based research (slides in English, as all the following): 31/10-03/11/2006
  3. Systems Dynamics and World Models — classical approaches to macro simulation, differential equations, macro simulation tools, qualitative simulation: 07-14/11/2006
  4. Microanalytical Simulation Models — classical approaches to micro simulation, tax and pension models, recent tools: 14-17/11/2006
  5. Queuing Models — discrete event simulation, business process modeling, tools: 21/11/2006
  6. Multilevel Modelling — modelling global interactions between populations, groups and individuals, stochastic processes, synergetics: 24-27/11-05/12/2006
  7. Cellular Automata — game theory, modelling local interactions in large populations of identical actors: 05-12/12/2006
  8. Distributed Artificial Intelligence Models — agent based social simulation: 15-19/12/2006
  9. Learning and Evolutionary Models — artificial neural networks, genetic algorithms: 09/01/2006

Additional models will be presented in the lectures on 12/01, 16/01, 19/01 and 13/02/2007.

Teaching material

All slides and a list of references will be available on the Internet. All models discussed during the lecture course (and some more) can be found in a WinZip archive. NetLogo can be downloaded from this site.

Student presentations

Groups of two or three students will present reports of their experience with simulation tools and simulation models during the last 8 meetings (23/01, 26/01, 30/01, 02/02, 06/02, 09/02, 16/02). Per meeting either two groups of three or three groups of two students will present. Choices can be made from System Dynamics, microsimulation, queuing Models, cellular automata and agent-based models. For microsiulation and queuing models, UMDBS and SimProcess, respectively, can be used (SimProcess is available on request), for the other approaches NetLogo can be used. Models can be found in the demo sections of the tools or in the WinZip archive mentioned above, and advice is given upon request.

23/01/2007: Queueing Models (Christopher Felix Wahl, Jan Wölker: Airport models in SimProcess and AnyLogic)

26/01/2007: Cellular automata (Vera Wolber, Ina Kimmling, Sebastian Wolf: Introduction, AIDS, CaFun, Termites)

30/01/2007: Systems Dynamics (Daniel Bender, Andreas Dapprich, Marc Drewing: Lotka-Volterra, World II) / Traffic Simulation (Valeska Kroheck, Waldemar Bergen: Car Traffic Flow)

02/02/2007: Cellular Automata (Patrick Faruhn, Daniel Mies, Michael Dornauf: RumorMill, Gossip and Epidemics);

06/02/2007: Microsimulation (Michael Zaggl, Bernd Anhäuser: Simulating the Effects of Parents' Subsidies); Cellular Automata (Anastasios Psarros, Matthias Joswig, Manuel Wagner, Onur Cakmak: Ethnocentricity, Wealth Distribution)

09/02/2007: Cellular Automata (Daniel Reiser, Christoph Schneider: Simulating a Product Market)

16/02/2007: Cellular Automata (Patrick Schober, Gernot Peters, David Dose, Monika Dammann: History, Bushfire, Co-operation, Innovation Management, Outlook: Irregular CAs))

Basic Reading

...