You might be interested in reading a portion of my book chapter:
"Community Viz: An Integrated Planning Support System" by Michael Kwartler and Robert N. Bernard, which is in the book "Planning Support Systems: Integrating GIS, Models, and Visualization Tools", edited by Richard Brail and Dick Klosterman, ESRI Press, Redlands, 2001. It's out of print now, but you can find it here: https://static1.squarespace.com/static/54d8cc78e4b003ad1dc6a0f7/54e377fae4b053bd0c641d50/54e377fae4b053bd0c641f25/1378331642313/Planning+Support+Systems.pdf )
It was written for a high-level audience, so it doesn't get into super-specific details, but it does discuss how we combined genetic algorithms and simulated annealing to calibrate a high-dimensional agent based simulation model (nearly 18 years ago -- how time flies)!
Rob Bernard
[log in to unmask]
-----Original Message-----
From: News and discussion about computer simulation in the social sciences <[log in to unmask]> On Behalf Of Tesfatsion, Leigh S [ECONS]
Sent: Monday, January 14, 2019 12:28
To: [log in to unmask]
Subject: Re: [SIMSOC] A review of optimisation/calibration methods for ABMs?
14 January 2019
Nick ---
cc: SimSoc
You might want to browse the annotated pointers to resources listed at the following site:
========================
Empirical Validation and Verification of Agent-Based Models http://www2.econ.iastate.edu/tesfatsi/EmpValid.htm
<http://www2.econ.iastate.edu/tesfatsi/EmpValid.htm>
Table of Contents:
Basic Issues
Introductory Materials on Empirical Validation
Introductory Materials on Verification
Additional Readings on Empirical Validation and Verification
Design of Computational Experiments
Iterative Participatory Modeling (IPM)
=======================
Suggestions for additional relevant resources to add at this site, from you or from other SimSoc participants, would be most welcome.
Best wishes,
Leigh
On 1/14/2019 10:19 AM, Nicolas Malleson wrote:
> Hi all,
>
> I am wondering if anyone is aware of a review or discussion about the quantitative methods used to calibrate/parameterise/optimise* agent-based models? E.g. discussing methods like genetic algorithms, regression, simulated annealing, etc., etc. with a view to recommending those that are the most suitable for different kinds of ABMs.
>
> To be honest I’m surprised not to have found anything, the need to estimate ‘optimal’ parameter values seems fairly common.
>
> Thanks in advance,
> Nick
>
>
> * I think that ‘calibrate', ‘parameterise’ and ‘optimise’ mean largely the same thing (or at least the same methods would be used).
>
> —
> Dr Nick Malleson
> Associate Professor in Geographical Information Science School of
> Geography, University of Leeds, LS2 9JT [log in to unmask]
> http://nickmalleson.co.uk/
> 0113 34 35248
>
>
> ######################################################################
> ##
>
> To unsubscribe from the SIMSOC list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=SIMSOC&A=1
--
Leigh Tesfatsion
Research Professor, and
Professor Emerita of Econ, Math, and
Electrical & Computer Engineering
Department of Economics/Heady 260
Iowa State University
Ames, Iowa 50011-1054
http://www2.econ.iastate.edu/tesfatsi/
[log in to unmask]
########################################################################
To unsubscribe from the SIMSOC list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=SIMSOC&A=1
########################################################################
To unsubscribe from the SIMSOC list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=SIMSOC&A=1
|