Framework for agent behavior and interactions: The BNAS model utilizes BNs to generate agent decisions through several model iterations (five years each).
The BNAS model then will select only related variables that will be used in the model.
The BNAS model allows the definition of dynamic variables (time-dependent drivers) at the start of each simulation.
The system of agents' iterations: The BNAS model is composed of static and variable parts.
Each agent has its own characteristics and field of vision and, in combination, they form a heterogeneous agent population in the BNAS model.
Given that the BNAS model is a dynamic model, the system runs each iteration 100 times.
The BNAS model does not define permanent data variables (such as land-use drivers).
In this study, the utility of the BNAS model in a PSS framework is tested by using it as a policy generator and a future land-use simulator.
The BNAS model can be employed in a PSS framework for processing information.
The BNAS model can incorporate land-use decisions through the links between the various drivers.
Figure 2 provides examples of BN structures for the household agents to illustrate how a BN in an agent type can be represented (these are examples and do not represent the BNAS model's variables).
This study expands on the initial BNAS model by using it as a policy generator through the backward inference capabilities of BNs.