ABMCSS Workshop Notes, 15 April 2007
Sat, 04/14/2007 - 23:50 — glennon
J. Alan Glennon
Department of Geography
University of California
Santa Barbara, CA 93106
email: alan at geog.ucsb.edu
Workshop webpage is:
8.30: Introductions, objectives of the workshop, logistics
Goodchild marriage of two projects:
U.S. -- computation and representation/dynamics/complexity/primitives
U.K. -- agent based modeling
Forty-nine workshop participants.
9:15AM--Mike Worboys, University of Maine
Modeling of Complex Spatial Systems
ontological support (time,change,events, processes...
for example, SNAP/SPAN, -continuant (perdurant) and snapshot
one approach: cellular automata
formalizing distributed processes
With a sensor, a location is a process.
Distributed processes are inherently complex, even when described with simple rules. High degree of uncertainty. Process algebra would be a next step beyond cellular automata rules.
10:30AM-- Marina Alberti, University of Washington
Modeling Challenges: emergent properties, adaptive agents, multiple equilibria, hierarchies, heterogeneity, path-dependency.
11:15AM--Nigel Gilbert, University of Surrey, discussant
I. Distinction between abstract models and facsimile models
II. Emergence. how to approach process algebra with respect to "emergence." Connecting mechanisms and activities.
1:30PM--David Bennett, University of Iowa
Agent-based modeling with context/location-aware decisionmaking
individual: perceive, plan, act, learn, adapt
context: environment, individual state, socioeconomic setting
representing individual cognition: motives, knowledge, capabilities, beliefs, certainty, social network.
2:15PM--Giorgio Theodoropoulos, University of Birmingham, discussant
Dynamic data driven application systems
-a simulation coexists and runs in parallel with the real system.
-feed from real-world sensors to simulation.
using DDDAS in social science-->requires ontologies/semantic descriptions of social systems.
prototype is AIMSS (housing happiness)
3:30PM--May Yuan, University of Oklahoma
Temporal GIS for agent-based modeling of complex spatial systems
Going back and forth from temporal GIS and agent-based modeling
drivers: activities, events, and processes
observables: change and movement
Hierarchy theory -- definitional and empirical entities.
Agent versus GIS data objects
data objects->empirical objects->inform definitional objects->agents
agents->definitional object->inform observations->data objects
4:40PM--David Maguire and Kevin Johnston, ESRI
Dynamic modeling with geo-agents
modeling in ArcGIS and Agent Analyst
ABMCSS Workshop Notes, 16 April 2007
Mon, 04/16/2007 - 09:19 — glennon
Grounding the debate about calibration and validation of spatial agent-based models
10:30AM--Michael Batty, University College London, discussant