Theme 5
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View Theme 5 as PDF for free.
More details
- Words: 1,069
- Pages: 3
Theme 5: Cellular, automata, agent-based modelling and simulation David F. Batten, Coordinator, CSIRO Agent-Based Modelling (CABM) Working Group Since the release of John Conway’s Game of Life, cellular automata have been used as models in many areas of the physical and spatial sciences, biology, mathematics and computer science – as well as in the social sciences. They are a useful modelling platform, as cells on a grid that switch on or off according to states of neighbouring cells can represent a host of dynamic phenomena – individuals, attitudes or actions, for example. A cellular automaton (CA) models any world in which space can be represented as a uniform grid, time advances by steps, and the “laws” of that world are represented by a uniform set of rules which compute each cell’s state from its own previous state and those of its nearby neighbours.
Once we wish to develop automata that are more complex in their internal processing and in their behaviour, we enter a different world known as agent-based modelling, multiagent systems or simulation. Such automata are called agents, and there are several streams of thought on how the agents can or should be designed, built and used. While there is no accepted definition of what an agent is, the term often signifies an autonomous, intelligent entity that can interact or communicate with other autonomous, intelligent entities. As in CA, there are rules governing interactive behaviour and the agents “operate” in or on an environment of some sort. The agents emanating from the literature on (distributed) artificial intelligence often correspond to self-contained software and/or hardware (e.g. robots) that control their own actions based on perceptions of their operating environments. Multiple agents are designed to work together to achieve a desired goal. Typically, their goal is pre-specified and they are engineered and controlled to achieve it with very little tolerance for error. Although purposive, systems in which human agents interact with ecological systems, for example, display open-ended outcomes. Here the collective behaviour is unknown in advance, but emerges during the simulation. Some emergent outcomes may be unexpected and undesirable. Artificial life has been the source of inspiration for this open-ended kind of simulation. In agent-based simulation, rules governing agents’ behaviours can range from simple “IF—THEN” statements to sophisticated machine learning algorithms (such as genetic algorithms) that allow agents to modify and improve their behaviour during the simulation. Data mining is often used to ensure that agents behave in ways that realistically depict how individual decisions are made in that system. Parameters of the model are set to represent a situation of interest and the model is run for several hundreds of iterations, until a preferred solution is found. Where possible, simulation models are calibrated against historical data to ensure that the model is accurately replicating the qualitative behaviour of the real system.
Agent-based simulations can provide valuable information about the dynamics of the real world(s) that they emulate. Complex systems scientists see them as more useful than equations-based methods in a world of multiple possible futures, partly because they are built synthetically “from the bottom up”. As a wide variety of agents interact within a social simulation, for example, results show how their collective behaviours govern the performance of the entire system – for instance, the emergence of a successful product, a congested area of traffic, or a polluted water catchment. This is of great benefit to stakeholders, because they can see a role for themselves (as agents) in the simulation itself, as well as an opportunity to learn from the simulated outcomes. Thus one area of the agent-based modelling field that is growing rapidly is known as companion modelling – the use of participatory agent-based simulation to improve our understanding of human (stakeholder) interactions with an environment. These simulations are particularly valuable for natural resource management. Furthermore, agent-based simulation is a powerful tool for “What if” scenario analysis. As some agents’ traits or behavioural rules change, the impact of the change can be seen in the model’s collective output. It is these adaptive learning features that give agentbased simulation an edge over more traditional modelling and optimisation methods. More importantly, the computer can generate outcomes or strategies that a scientist or stakeholder might never have
imagined. Some might say that this new kind of science is helping us to unravel the complexity of the world around us and deepen our understanding of how many human creations might grow and change over time. Observation and experimentation are still alive and well, but have been joined by an entirely new breed of computational science: simulation. We can make a strong argument that the sort of science we do when we simulate – particularly when we simulate complex adaptive systems such as societies, organisms or the human brain – is sufficiently different to normative science that we may call it a new kind of science. Its “new-kindedness” allows it to unravel some of the organized complexity around us, and to understand interdependencies between the slower and faster processes of change that affect us all. Agent-based modelling and simulation are at the forefront of this new kind of science. Studying a system that possesses a large number of adaptive agents, with intricate feedback loops linking macrostructure and microbehaviour, usually rules out an equations-based model. We must resort to computer simulation. The fundamental class of properties of the social world that this kind of simulation opens to new understanding is that which occurs only in the dynamics produced by the accumulated interactions of the agents making up the complex adaptive system (CAS). Some Background References: Batty, M. (2005): Cities and Complexity: Understanding Cities with Cellular Automata,
Agent-Based Models and Fractals, Cambridge: The MIT Press. Casti, J. (1997): Would-Be Worlds: How Simulation is Changing the Frontiers of Science, New York: Wiley. Epstein, J. and R. Axtell (1995): Growing Artificial Societies: Social Science from the Bottom Up, Washington: Brookings Institution Press. Ferber, J. (1999): Multi-Agent Systems: An Introduction to Distributed Artificial Intelligence. New York: Addison Wesley. Janssen, M., ed. (2002): Complexity and Ecosystem Management: The Theory and Practice of Multi-agent Systems. Cheltenham: Edward Elgar. Perez, P and D. Batten, eds (2006): Complex Science for a Complex World: Exploring Human Ecosystems with Agents, Canberra: ANU ePress. (http://epress.anu.edu.au). Tesfatsion, L and K. Judd, eds. (2006): Handbook of Computational Economics: Volume 2 – Agent-based Computational Economics, Amsterdam: North-Holland. Wolfram, Stephen (2002): A New Kind of Science, Champaign: Wolfram Science.
Once we wish to develop automata that are more complex in their internal processing and in their behaviour, we enter a different world known as agent-based modelling, multiagent systems or simulation. Such automata are called agents, and there are several streams of thought on how the agents can or should be designed, built and used. While there is no accepted definition of what an agent is, the term often signifies an autonomous, intelligent entity that can interact or communicate with other autonomous, intelligent entities. As in CA, there are rules governing interactive behaviour and the agents “operate” in or on an environment of some sort. The agents emanating from the literature on (distributed) artificial intelligence often correspond to self-contained software and/or hardware (e.g. robots) that control their own actions based on perceptions of their operating environments. Multiple agents are designed to work together to achieve a desired goal. Typically, their goal is pre-specified and they are engineered and controlled to achieve it with very little tolerance for error. Although purposive, systems in which human agents interact with ecological systems, for example, display open-ended outcomes. Here the collective behaviour is unknown in advance, but emerges during the simulation. Some emergent outcomes may be unexpected and undesirable. Artificial life has been the source of inspiration for this open-ended kind of simulation. In agent-based simulation, rules governing agents’ behaviours can range from simple “IF—THEN” statements to sophisticated machine learning algorithms (such as genetic algorithms) that allow agents to modify and improve their behaviour during the simulation. Data mining is often used to ensure that agents behave in ways that realistically depict how individual decisions are made in that system. Parameters of the model are set to represent a situation of interest and the model is run for several hundreds of iterations, until a preferred solution is found. Where possible, simulation models are calibrated against historical data to ensure that the model is accurately replicating the qualitative behaviour of the real system.
Agent-based simulations can provide valuable information about the dynamics of the real world(s) that they emulate. Complex systems scientists see them as more useful than equations-based methods in a world of multiple possible futures, partly because they are built synthetically “from the bottom up”. As a wide variety of agents interact within a social simulation, for example, results show how their collective behaviours govern the performance of the entire system – for instance, the emergence of a successful product, a congested area of traffic, or a polluted water catchment. This is of great benefit to stakeholders, because they can see a role for themselves (as agents) in the simulation itself, as well as an opportunity to learn from the simulated outcomes. Thus one area of the agent-based modelling field that is growing rapidly is known as companion modelling – the use of participatory agent-based simulation to improve our understanding of human (stakeholder) interactions with an environment. These simulations are particularly valuable for natural resource management. Furthermore, agent-based simulation is a powerful tool for “What if” scenario analysis. As some agents’ traits or behavioural rules change, the impact of the change can be seen in the model’s collective output. It is these adaptive learning features that give agentbased simulation an edge over more traditional modelling and optimisation methods. More importantly, the computer can generate outcomes or strategies that a scientist or stakeholder might never have
imagined. Some might say that this new kind of science is helping us to unravel the complexity of the world around us and deepen our understanding of how many human creations might grow and change over time. Observation and experimentation are still alive and well, but have been joined by an entirely new breed of computational science: simulation. We can make a strong argument that the sort of science we do when we simulate – particularly when we simulate complex adaptive systems such as societies, organisms or the human brain – is sufficiently different to normative science that we may call it a new kind of science. Its “new-kindedness” allows it to unravel some of the organized complexity around us, and to understand interdependencies between the slower and faster processes of change that affect us all. Agent-based modelling and simulation are at the forefront of this new kind of science. Studying a system that possesses a large number of adaptive agents, with intricate feedback loops linking macrostructure and microbehaviour, usually rules out an equations-based model. We must resort to computer simulation. The fundamental class of properties of the social world that this kind of simulation opens to new understanding is that which occurs only in the dynamics produced by the accumulated interactions of the agents making up the complex adaptive system (CAS). Some Background References: Batty, M. (2005): Cities and Complexity: Understanding Cities with Cellular Automata,
Agent-Based Models and Fractals, Cambridge: The MIT Press. Casti, J. (1997): Would-Be Worlds: How Simulation is Changing the Frontiers of Science, New York: Wiley. Epstein, J. and R. Axtell (1995): Growing Artificial Societies: Social Science from the Bottom Up, Washington: Brookings Institution Press. Ferber, J. (1999): Multi-Agent Systems: An Introduction to Distributed Artificial Intelligence. New York: Addison Wesley. Janssen, M., ed. (2002): Complexity and Ecosystem Management: The Theory and Practice of Multi-agent Systems. Cheltenham: Edward Elgar. Perez, P and D. Batten, eds (2006): Complex Science for a Complex World: Exploring Human Ecosystems with Agents, Canberra: ANU ePress. (http://epress.anu.edu.au). Tesfatsion, L and K. Judd, eds. (2006): Handbook of Computational Economics: Volume 2 – Agent-based Computational Economics, Amsterdam: North-Holland. Wolfram, Stephen (2002): A New Kind of Science, Champaign: Wolfram Science.
