Agent-Based Modeling
10 Nov 2024 19:01
Draft from circa 2004
Fundamentally, I'm not sure that agent-based modeling amounts to anything other than object-oriented programming for disaggregated simulations --- which is a very useful thing, of course. (I've expanded on this point in my review chapter on methods and techniques of complex systems theory.) I don't do much in this line, but it's important enough in the general area I work in that I feel like I ought to keep track of developments. This notebook is for methodology, not for particular substantive studies which happen to use agent-based models, unless they're exemplary in some way.
Has any work at all been done on statistical inference for agent-based models?
Update, 18 March 2007
A propos of the "agent-based == object-oriented" question, Peter McBurney writes:While object-oriented programming techniques can be used to design and build sofware agent systems, the technologies are fundamentally different. Software objects are encapsulated (and usually named) pieces of software code. Software agents are software objects with, additionally, some degree of control over their own state and their own execution. Thus, software objects are fixed, always execute when invoked, always execute as predicted, and have static relationships with one another. Software agents are dynamic, are requested (not invoked), may not necessarily execute when requested, may not execute as predicted, and may not have fixed relationships with one another.
Update, 13 April 2022
There is still not a huge amount on statistical inference for agent-based models, but see references below. In particular, I now have an idea which I'm very excited to try to get working here, namely random feature matching.- See also:
- Calculating Macroscopic Consequences of Microscopic Interactions
- Compartment Models
- Complexity
- Computational Models of Linguistic Evolution
- Ecology
- Economics
- Evolving Local Rules to Perform Global Computations
- Evolutionary Economics
- Flocking and Swarms
- Indirect Inference
- Interacting Particle Systems
- Learning in Games
- Macroscopic Consequences of Microscopic Interactions
- Multi-Agent Systems
- Schelling model
- Sociology
- Statistical Emulators for Simulation Models
- Recommended, big picture:
- Joshua M. Epstein, "Agent-based Computational Models and Generative Social Science", Complexity 4:5 (1999): 41--60 [Reprinted as chapter 1 of his Generative Social Science, not recommended simply because I haven't read all of it]
- Joshua M. Epstein and Robert Axtell, Growing Artificial Societies: Social Science from the Bottom Up
- Volker Grimm, Eloy Revilla, Uta Berger, Florian Jeltsch, Wolf M. Mooij, Steven F. Railsback, Hans-Hermann Thulke, Jacob Weiner, Thorsten Wiegand, and Donald L. DeAngelis, "Pattern-Oriented Modeling of Agent-Based Complex Systems: Lessons from Ecology", Science 310 (2005): 987--991
- Michael W. Macy and Robert Willer, "From Factors to Actors: Computational Sociology and Agent-Based Modeling," Annual Review of Sociology 2002
- John H. Miller and Scott E. Page, Complex Adaptive Systems: An Introduction to Computational Models of Social Life [Disclaimer: John is a friend and Scott is a former boss; both have been mentors. As a trivial show of independence, let me just say that I hate the phrase "complex adaptive systems".]
- Mitchel Resnick, Turtles, Termites and Traffic Jams: Explorations in Massively Parallel Microworlds
- Recommended, close-ups:
- Eric Bonabeau, "From Classical Models of Morphogenesis to Agent-Based Models of Pattern Formation", Artificial Life 3 (1997): 191--211
- Giovanni Luca Ciampaglia, "A framework for the calibration of social simulation models", Advances in Complex Systems 16 (2013): 1350030, arxiv:1305.3842
- Martin Clarke and Einar Holm, "Microsimulation Methods in Spatial An alysis and Planning", Geografiska Annaler. Series B, Human Geography 69 (1987): 145--164 [JSTOR. A largely independent tradition from the one I was taught.]
- Steven N. Durlauf and H. Peyton Young (eds.), Social Dynamics
- H. Randy Gimblett (ed.), Integrating Geographic Information Systems and Agent-Based Modeling Techniques for Understanding Social and Ecological Processes
- Robert L. Goldstone and Marco A. Janssen, "Computational models of collective behavior", Trends in Cognitive Sciences 9 (2005): 424--430 [Brief review on agent-based models]
- Christian Gouriéroux and Alain Monfort, Simulation-Based Econometric Methods [Doesn't mention agent-based models at all, but the principles it describes for doing statistical inference on complicated simulation models would certainly apply. Review: By Indirection Find Direction Out.]
- Peter Hedstrom, Dissecting the Social: On the Principles of Analytical Sociology [See especially chapter 6, co-written with Yvonne Aberg. Some notes on that chapter, for an old class]
- Mevin B. Hooten and Christopher K. Wikle, "Statistical Agent-Based Models for Discrete Spatio-Temporal Systems", Journal of the American Statistical Association 105 (2010): 236--248
- Aki Lehtinen and Jaakko Kuorikoski, "Computing the Perfect Model: Why Do Economists Shun Simulation", Philosophy of Science 74 (2007): 304--329 [What they have to say sounds right, but more like a reason for economists to change their ideals than anything else.]
- Kristina Lerman, "Design and Mathematical Analysis of Agent-based Systems" [Preprint]
- Kristina Lerman, Aram Galstyan, Tad Hogg, "Mathematical Analysis of Multi-Agent Systems", arxiv:cs.MA/0404002
- John H. Miller, "Active Nonlinear Tests (ANTs) of Complex Simulation Models", Management Science 44 (1998): 820--830 [PDF reprint; thanks to Will Tracy for letting me know about this paper]
- Scott Moss and Bruce Edmonds, "Sociology and Simulation: Statistical and Qualitative Cross‐Validation", American Journal of Sociology 110 (2005): 1095--1131 [This is an interesting but uneven paper. The strong points are hammering home the possibility, and utility, of employing agent-based models of social situations which start from qualitative ideas about individual-level behavior and interactions, and checking the simulation output against aggregated real data. They're also quite right that strong interactions and meta-stability are apt to generate clustered volatility and (therefore) heavy tails in aggregate time series. There are also some fair strictures against just doing regressions, or just employing clustered volatility models that don't actually explain anything. There is an implication (not fully explicit) that ideally social-science models should be generative, which I endorse. But: like many sociological papers, it's got lots of little digs against economists, which are not always fair or accurate, and it's got some weird ideas about statistics. Thus, in contradiction to the authors: Agents not altering their behavior in response to small stimuli (like a penny change in the price of a can of tuna) is fully compatible with agents being utility maximizing (control theory teaches us that maximizing continuous objectives in continuous time can sometimes best be done by "stop-go" action); ordinary least squares linear regression is fully compatible with very heavy tails in the noise terms, though that screws up the conventional formulas for confidence intervals and hypothesis tests (but you should be using the bootstrap for small samples anyway); saying that Pareto distributions don't have a well-defined mean over half their parameter space is just weird (I can come up with a parameterization where it's true, but it's weird and non-standard, in the usual one they have well-defined variance for most exponents!); chi-squared tests are asymptotically valid for lots of non-Gaussian situations (e.g., when using maximum likelihood estimates); their agent-based model is a parametric statistical model; most annoyingly, "cross-validation" had been an established statistical term since the 1970s, in a very different sense than the authors use it. (Roughly speaking, M&E call a model "cross-validated" when there are distinct sources of evidence which all support it, whereas statistical cross-validation is seeing how well the model can predict data it wasn't estimated on. Their "cross-validation" is actually pretty close to E. O. Wilson's "consilience".) I'd be happy to teach this paper in a class on the statistics of ABMs, but I'd have to give my students a very carefully annotated copy.]
- Karl Naumann-Woleske, Max Sina Knicker, Michael Benzaquen, Jean-Philippe Bouchaud, "Exploration of the Parameter Space in Macroeconomic Agent-Based Models", arxiv:2111.08654 [Thanks to Dr. Bouchaud for alerting me to this paper]
- J.-H. Niemann, S. Winkelmann, S. Wolf, C. Schütte, "Agent-based modeling: Population limits and large timescales", Chaos 31 (2021): 033140
- Paul Windrum, Giorgio Fagiolo and Alessio Moneta, "Empirical Validation of Agent-Based Models: Alternatives and Prospects", Journal of Artificial Societies and Social Simulation 10:2 (2007): 8
- Recommended, inspirations:
- Robert Axelrod, The Evolution of Cooperation [A major inspiration, especially for people working in the social sciences]
- Christopher Langton (ed.), Artificial Life [Along with subsequent volumes in the series; many of the papers here and later are not however particularly relevant to this theme]
- Francisco J. Varela and Paul Bourgine (eds.), Toward a Practice of Autonomous Systems: Proceedings of the First European Conference on Artificial Life
- Modesty forbids me to recommend:
- CRS, "Methods and Techniques of Complex Systems Science: An Overview", pp. 33--114 in Thomas S. Deisboeck and J. Yasha Kresh (eds.), Complex Systems Science in Biomedicine, arxiv:nlin.AO/0307015, especially section 5
- CRS, "A Note on Simulation-Based Inference by Matching Random Features", arxiv:2111.09220 [auto-vulgarization]
- To read:
- Alberto Acerbi, Alex Mesoudi, Marco Smolla, Individual-based models of cultural evolution: A step-by-step guide using R
- Christoph Adami, Jory Schossau, Arend Hintze, "Evolutionary game theory using agent-based methods", arxiv:1404.0994
- Masanao Aoki
- Sven Banisch, Ricardo Lima, Tanya Araújo
- "Agent Based Models and Opinion Dynamics as Markov Chains", arxiv:1108.1716
- "Aggregation and Emergence in Agent-Based Models: A Markov Chain Approach", arxiv:1207.2255
- Jason Barr, Troy Tassier and Leanne Ussher (eds.), Symposium on Agent-Based Computational Economics, special issue (37:1 [Winter 2011]) of Eastern economic Journal
- Garrett Bernstein, Kyle O'Brien, "Stochastic Agent-Based Simulations of Social Networks", arxiv:1309.1747
- Elizabeth Bruch and Jon Atwell, "Agent-Based Models in Empirical Social Research", Sociological Methods & Research 44 (2015): 186--221
- Ernesto Carrella, "No Free Lunch when Estimating Simulation Parameters", Journal of Artificial Societies and Social Simulation 24:2 (2021): 7
- Anirban Chakraborti, Damien Challet, Arnab Chatterjee, Matteo Marsili, Yi-Cheng Zhang, Bikas K. Chakrabarti, "Statistical Mechanics of Competitive Resource Allocation using Agent-based Models", arxiv:1305.2121
- Alan Dorin and Nicholas Geard, "The Practice of Agent-Based Model Visualization", Artificial Life 20 (2014): 271--289
- Igor Douven, "Social Learning in Neural Agent-Based Models", Philosophy of Science fortchoming
- Joel Dyer, Patrick Cannon, J. Doyne Farmer, Sebastian M. Schmon, "Calibrating Agent-based Models to Microdata with Graph Neural Networks", arxiv:2206.07570
- Joshua M. Epstein
- Radek Erban, Ioannis G. Kevrekidis and Hans G. Othmer, "An equation-free computational approach for extracting population-level behavior from individual-based models of biological dispersal", physics/0505179
- Daniel Frey and Dunja Seselja, "Robustness and Idealizations in Agent-Based Models of Scientific Interaction", The British Journal for the Philosophy of Science 71 (2020): 1411--1437
- Heather J. Goldsby, Anna Dornhaus, Benjamin Kerr, and Charles Ofria, "Task-switching costs promote the evolution of division of labor and shifts in individuality", Proceedings of the National Academy of Sciences (USA) 109 (2012): 13686--13691
- Miguel Gonzalez, Richard Watson, Seth Bullock, "Minimally Sufficient Conditions for the Evolution of Social Learning and the Emergence of Non-Genetic Evolutionary Systems", Artificial Life 23 (2017): 493--517
- Volker Grimm and Steven F. Railsback, Individual-based Modeling and Ecology
- Till Gruene-Yanoff, "Agent-Based Simulation, Generative Science, and Its Explanatory Claims", phil-sci/2784
- Daniel Heard, Gelonia Dent, Tracy Schifeling, and David Banks, "Agent-Based Models and Microsimulation", Annual Review of Statistics and Its Application 2 (2015): 259--272
- Peter Hedström and Gianluca Manzo (eds.), Agent-Based Modeling: Advances and Challenges, special issue of Sociological Methods and Research 44:2 (2015) [editors' introduction]
- Franziska Hinkelmann, David Murrugarra, Abdul Salam Jarrah, Reinhard Laubenbacher, "A Mathematical Framework for Agent Based Models of Complex Biological Networks", arxiv:1006.0408
- Jack D. Hywood, Emily J. Hackett-Jones, and Kerry A. Landman, "Modeling biological tissue growth: Discrete to continuum representations", Physical Review E 88 (2013): 032704
- Bjorn Erik Juel, Renzo Comolatti, Giulio Tononi, Larissa Albantakis, "When is an action caused from within? Quantifying the causal chain leading to actions in simulated agents", arxiv:1904.02995
- Nicholas Lanicher, "The Axelrod model for the dissemination of culture revisited", Annals of Applied Probability 22 (2012): 860--880
- Roberto Leombruni and Matteo Richiardi, "Why are economists sceptical about agent-based simulations?", Physica A 355 (2005): 103--109 ["We look at the following problematic areas: (i) interpretation of the simulation dynamics and generalization of the results, and (ii) estimation of the simulation model. We show that there exist solutions for both these issues."]
- Laura Liu, Mikkel Plagborg-Møller, "Full-Information Estimation of Heterogeneous Agent Models Using Macro and Micro Data", arxiv:2101.04771
- Nadia Loy, Andrea Tosin, "Markov jump processes and collision-like models in the kinetic description of multi-agent systems", Communications in Mathematical Sciences 18 (2020): 1539--1568, arxiv:1905.11343
- Michael Luck, Peter McBurney, Onn Shehory, Steve Willmott, Agent Technology: Computing as Interaction (A Road-Map for Agent Based Computing) [Available as a PDF from AgentLink. Thanks to Prof. McBurney for letting me know about this]
- Daniel McDuff, Yale Song, Jiyoung Lee, Vibhav Vineet, Sai Vemprala, Nicholas Gyde, Hadi Salman, Shuang Ma, Kwanghoon Sohn, Ashish Kapoor, "CausalCity: Complex Simulations with Agency for Causal Discovery and Reasoning", arxiv:2106.13364
- Ruben Mercado, Artificial Economics: Methods, Models, and Interdisciplinary Links
- Corrado Monti, Marco Pangallo, Gianmarco De Francisci Morales and Francesco Bonchi, "On learning agent-based models from data", Scientific Reports 13 (2023): 9268
- Akira Namatame, Shu-Heng Chen, Agent-Based Modeling and Network Dynamics
- Michael J. North and Charles M. Macal, Managing Business Complexity: Discovering Strategic Solutions with Agent-Based Modeling and Simulation
- Cathal O'Donoghue, Practical Microsimulation Modelling
- Catherine J. Penington, Barry D. Hughes, and Kerry A. Landman, "Building macroscale models from microscale probabilistic models: A general probabilistic approach for nonlinear diffusion and multispecies phenomena", Physical Review E 84 (2011): 041120
- Sebastian Poledna, Michael Gregor Miess and Cars Hommes, "Economic Forecasting with an Agent-based Model" [Preprint]
- Wolfgang Radax, Bernhard Rengs, "Timing matters: Lessons From The CA Literature On Updating", arxiv:1008.0941
- Steven Railsback and Volker Grimm, Agent-Based and Individual-Based Modeling: A Practical Introduction
- Steven F. Railsback and Bret C. Harvey, Modeling Populations of Adaptive Individuals
- Steven F. Railsback, Steven L. Lytinen and Stephen K. Jackson, "Agent-based Simulation Platforms: Review and Development Recommendations", Simulation 82 (2006): 609--623
- Daniel Remenik, "Limit Theorems for Individual-Based Models in Economics and Finance", Stochastic Processes and their Applications 119 (2009): 2401--2435, arxiv:0810.2813
- Tomas Salamon, Design of Agent-Based Models: Developing Computer Simulations for a Better Understanding of Social Processes
- Javier Segovia, "Statistical Thermodynamics Concepts and Mathematical Tools for a Multi-Agent Ecosystem", Artificial Life 20 (2014): 237--270
- Karandeep Singh, Chang-Won Ahn, Euihyun Paik, Jang Won Bae, Chun-Hee Lee, "A Micro-Level Data-Calibrated Agent-Based Model: The Synergy between Microsimulation and Agent-Based Modeling", Artificial Life 24 (2018): 128--148
- Alexander Smajgl and Olivier Barreteau (eds.), Empirical Agent-Based Modelling --- Challenges and Solutions, Volume 1, The Characterisation and Parameterisation of Empirical Agent-Based Models
- Paul E. Smaldino, Modeling Social Behavior: Mathematical and Agent-Based Models of Social Dynamics and Cultural Evolution
- Leon Sterling and Kuldar Taveter, The Art of Agent-Oriented Modeling [Software engineering]
- Leigh Tesfatsion and Kenneth L. Judd (eds.), Agent-Based Computational Economics, vol. 2 of the Handbook of Computational Economics
- Thomas Thelen, Gerardo Aldana, Marcus Thomson, Toni Gonzalez, "villager: A Framework for Designing and Running Agent Based Models", CRAN
- Jan C. Thiele and Volker Grimm, "NetLogo meets R: Linking agent-based models with a toolbox for their analysis", Environmental Modelling and Software 25 (2010): 972--974
- Jan C. Thiele, Winfried Kurth and Volker Grimm, "Agent-Based Modelling: Tools for Linking NetLogo and R", Journal of Artificial Societies and Social Simulation 15:3 (2012): 8
- Thomas N. Thiem, Felix P. Kemeth, Tom Bertalan, Carlo R. Laing and Ioannis G. Kevrekidis, "Global and Local Reduced Models for Interacting, Heterogeneous Agents", Chaos 31 (2021): 073139, arxiv:2105.09398
- D. Townsend, "Validation and Inference of Agent Based Models", arxiv:2107.03619
- M. Utku Unver, "Backward unraveling over time: The evolution of strategic behaivor in the entry-level British medical labor markets", Journal of Economic Dynamics and Control 25 (2001): 1039--1080 [Thanks to Will Tracy for a copy]
- Konstantina Valogianni and Balaji Padmanabhan, "Causal ABMs: Learning Plausible Causal Models using Agent-based Modeling", pp. 3--29 in Thuc Duy Le, Lin Liu, Emre Kiciman, Sofia Triantafyllou and Huan Liu (eds.), proceedings of The KDD'22 Workshop on Causal Discovery
- Michael P. Wellman, "Putting the agent in agent-based modeling", Autonomous Agents and Mutli-Agent Systems 30 (2016): 1175--1189
- Uri Wilensky and William Rand, An Introduction to Agent-Based Modeling: Modeling Natural, Social, and Engineered Complex Systems with NetLogo
- Biao Wu, "Interacting Agent Feedback Finance Model", math.PR/0703827
- Petri Ylikoski, "Agent-Based Simulation and Sociological Understanding", Perspectives on Science 22 (2014): 318--335