Urban life: a model of people and places

The Ground Truth program: simulations as test beds for social science research methods

“…The complexity of the social simulation test beds varied over the challenges, reflecting a range of types of complexity (Castellani 2014), while each type of test assessed a different set of Three categories of teams participated in the program, including the social simulation teams, the research teams, and a test and evaluation team (which for this discussion includes program design/management and other government partners). Four simulation teams designed and built the simulation test beds (Parunak 2022;Pynadath et al 2022;Rager et al 2022;Züfle et al 2022). The simulations focused on different Human Domain topics and systems (urban life, financial governance, disaster response, and geopolitical conflict), but all of the simulations involved actors making decisions, implementing behaviors, and interacting with each other and their environments.…”

Simulating patterns of life: More representative time-activity patterns that account for context

“…Each individual is represented as an agent with a personal schedule; in combination with contextual factors, the behavior modeling involves simulating the interactions of very many agents. The intent is to identify emergent patterns from numerous runs of agents’ activities over some time period (e. g., a day or week), including their interactions with each other ( Jiang et al, 2020 ; Züfle et al, 2021 ) and their environment ( Abraham et al, 2017 ; Rand and Stummer, 2021 ; Zhu et al, 2013 ), based on imposed rules. For instance, many iterations of community-level agent-based models, covering diverse activities among individuals, yield usable patterns of culturally-aware agents that exhibit behaviors that reflect locally meaningful customs or that change based on local preferences or demands ( Bergier and Faucher, 2018 ; Tracy et al, 2018 ; Xu, 2012 ).…”

Methods That Support the Validation of Agent-Based Models: An Overview and Discussion

“…Examinations of the behaviors occurring between the agents and environment, i.e., the agent-environment considerations, are explorable through the use of pattern-based visualization methods, such as cluster heat maps(St-Aubin et al 2018;Wilkinson & Friendly 2009), and location tracking(Kim et al 2019;Züfle et al 2021). Cluster heat maps can yield insights into concentrations of agents based on the intersection of selected agent characteristics and the agents' associated geolocations at specific points in time(Diallo et al 2021;Lynch et al 2019).…”