Using Bayesian Decision Networks for Knowledge Representation under Conditions of Uncertainty in Multi-Agent Land Use Simulation Models

Ma, Linda; Arentze, Theo; Borgers, Awj Aloys; Timmermans, Harry

doi:10.1007/1-4020-2409-6_9

Cited by 9 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the late 1980s, researchers in artificial intelligence discovered that BNs provided an efficient means for handling uncertainty (Charniak, 1991;Heckerman et al, 1995;Pearl, 1988). Early applications of BNs were in medical diagnosis and genetics, but now their use can also be found in environmental studies (Little et al, 2004;Marcot et al, 2001;Varis, 1998), watershed management (Borsuk et al, 2001), remote sensing (Qin et al, 2006), land-use change (Kocabas and Dragicevic, 2006), land-use decision making (Ma et al, 2004), and urban land market sales (Lei et al, 2005).…”

Section: Bnsmentioning

confidence: 99%

Agent-based model validation using Bayesian networks and vector spatial data

Kocabas

Dragićević

2009

Environ. Plann. B

View full text Add to dashboard Cite

Validation of agent-based models (ABMs) of land-use change is a significant challenge in current spatial-modelling research and application. During the validation process, model performance and accuracy assessment depend mostly on pixel-by-pixel comparisons. However, in urban land-use planning problems the use of vector spatial data to develop ABMs is becoming more necessary. Hence, improved and robust validation approaches are required for vector-based ABMs. This study presents a novel validation approach for an ABM using vector-based geographic information system and Bayesian networks. The approach creates a unique-polygons map and an object-oriented database. Three indicator variables are calculated to assess the probability of agreement. The indicator variables are nodes in a Bayesian network that is used to evaluate the final agreement of each unique polygon. Further, an index of overall agreement is calculated. The approach was applied to a simulation outcome map generated by an existing Bayesian network-based agent-system (BNAS) model. The BNAS model simulation of land-use change for the year 2001 was compared with the actual land-use change for the same year using the proposed validation approach. The results obtained indicate that significant agreement between the maps was achieved. The approach is well suited for validating vector-based ABMs and can be used as an aid in model designs for improved model performance.

show abstract

Section: Bnsmentioning

confidence: 99%

Agent-based model validation using Bayesian networks and vector spatial data

Kocabas

Dragićević

2009

Environ. Plann. B

View full text Add to dashboard Cite

show abstract

“…An agent constructs his or her goal state belief in three consecutive steps. First, the agent applies a decision model in order to determine the expected utility U of each cell based on the information available (for a more detailed study concerning the decision model, see Ma et al, 2004). At the beginning of the process this information is limited to the attribute data found in the available information sources, but as the process progresses more accurate information becomes available as the initiator makes assignment decisions.…”

Section: Goal State Beliefmentioning

confidence: 99%

A Multiagent Model for Alternative Plan Generation

Saarloos

Arentze

Borgers

et al. 2005

Environ Plann B Plann Des

Self Cite

View full text Add to dashboard Cite

The utilization of multiagent technology for urban planning purposes has already received much attention with regard to predicting and evaluating the effects of different policy scenarios and plan alternatives. The generation of these plan alternatives, however, remains underexplored in this context. In this paper the authors describe a multiagent model for generating alternative land-use plans, in which the agents are land-use experts that initiate the development of plan proposals and communicate with each other over time in order to draw up the proposals incrementally. The authors present a probabilistic approach to agent behaviour to enable decisionmaking under conditions of uncertainty. They describe what personal and collective beliefs agents construct and use in order to choose their actions strategically. The negotiation, taking place between the initiating agent and the other agents, aimed at reaching agreement with regard to the various claims, is organized as an iterative process in which both parties consider conciliatory adjustments to their strategies, and thus their decisions, in order to try to find mutually acceptable solutions. A hypothetical case study demonstrates the operation of the model and the effects of its behavioural parameters.

show abstract

“…Several studies have combined BNs with ABMs; for example, Kocabase and Dragicevic (2013) derived BN structures for different agent types and obtained their CPT values from census data before using the BN in a land-use change model [5]. In another example, both Ren and Anumba [25] and Ma et al [26] used a simple BN structure, utilised experts to derive their CPT values, and trained during the simulation [19,27]. While Matsumoto et al (2017) constructed a data-driven BN using a survey, they simultaneously trained their network to estimate internal parameters [28].…”

Section: Introductionmentioning

confidence: 99%

Bayesian networks for spatial learning: a workflow on using limited survey data for intelligent learning in spatial agent-based models

et al. 2019

View full text Add to dashboard Cite

Machine learning (ML) algorithms steer agent decisions in agent-based models (ABMs), serving as a vehicle for implementing behaviour changes during simulation runs. However, when training an ML algorithm, obtaining large sets of micro-level human behaviour data is often problematic. Information on human behaviour is often collected via surveys of relatively small sample sizes. This paper presents a methodology for training a learning algorithm to guide agent behaviour in a spatial ABM using a limited survey data sample. We apply different implementation strategies using survey data and Bayesian networks (BNs). By being grounded in probabilistic directed graphical models, BNs stand out among other learning algorithms in that they can be based on expert knowledge and/or known datasets. This paper presents four alternative implementations of data-driven BNs to support agent decisions in a spatial ABM. We differentiate between training BNs prior to, or during the simulation runs, using only survey data or a combination of survey data and expert knowledge. The four different implementations are then illustrated using a spatial ABM of cholera diffusion for Kumasi, Ghana. The results indicate that a balance between expert knowledge and survey data provides the best control over the learning process of the agents and produces the most realistic agent behaviour.

show abstract

Using Bayesian Decision Networks for Knowledge Representation under Conditions of Uncertainty in Multi-Agent Land Use Simulation Models

Cited by 9 publications

References 10 publications

Agent-based model validation using Bayesian networks and vector spatial data

Agent-based model validation using Bayesian networks and vector spatial data

A Multiagent Model for Alternative Plan Generation

Bayesian networks for spatial learning: a workflow on using limited survey data for intelligent learning in spatial agent-based models

Contact Info

Product

Resources

About