Agent-based Modeling Research Articles

Investigating resilience in the childcare context through the agent-based paradigm.

Investigating resilience in the childcare context through the agent-based paradigm.

Bayesian Calibration of Stochastic Agent Based Model via Random Forest.

Agent-based models (ABM) provide an excellent framework for modeling outbreaks and interventions in epidemiology by explicitly accounting for diverse individual interactions and environments. However, these models are usually stochastic and highly parametrized, requiring precise calibration for predictive performance. When considering realistic numbers of agents and properly accounting for stochasticity, this high-dimensional calibration can be computationally prohibitive. This paper presents a random forest-based surrogate modeling technique to accelerate the evaluation of ABMs and demonstrates its use to calibrate an epidemiological ABM named CityCOVID via Markov chain Monte Carlo (MCMC). The technique is first outlined in the context of CityCOVID's quantities of interest, namely hospitalizations and deaths, by exploring dimensionality reduction via temporal decomposition with principal component analysis (PCA) and via sensitivity analysis. The calibration problem is then presented, and samples are generated to best match COVID-19 hospitalization and death numbers in Chicago from March to June in 2020. These results are compared with previous approximate Bayesian calibration (IMABC) results, and their predictive performance is analyzed, showing improved performance with a reduction in computation.

Advances in computational modeling of cytokine and growth factor dynamics in bone healing: a scoping review.

Bone healing is a complex process regulated by intricate biological and mechanical factors and spatially varied regions over time. This scoping review synthesizes current computational models that incorporate cytokines and growth factors, examining their role in bone healing. Through a systematic analysis of 71 studies, this review identifies and categorizes the modeling methodologies used, including mathematical, finite element, agent-based, mechanobiological, pharmacobiological, and hybrid approaches. The findings highlight the predominant use of mathematical models while noting a recent shift toward more sophisticated techniques like finite element and agent-based models. Key cytokines and growth factors, such as TGF-β, RANK-RANKL-OPG, and PTH, are repeatedly used, underscoring their essential roles in regulating cellular processes. This review also analyzes parameter selection and validation strategies, identifying gaps in current practices and emphasizing the need for high-quality experimental validation to improve model reliability. Some bibliometric analyses provide insights into citation networks and keyword co-occurrence, illustrating influential studies in the field and central themes. The findings offer a foundation for future research to enhance model accuracy, aiming toward more predictive and clinically relevant models accounting for biology and mechanics in bone healing.

Control of medical digital twins with artificial neural networks.

The objective of precision medicine is to tailor interventions to an individual patient's unique characteristics. A key technology for this purpose involves medical digital twins, computational models of human biology that can be personalized and dynamically updated to incorporate patient-specific data. Certain aspects of human biology, such as the immune system, are not easily captured with physics-based models, such as differential equations. Instead, they are often multi-scale, stochastic and hybrid. This poses a challenge to existing control and optimization approaches that cannot be readily applied to such models. Recent advances in neural-network control methods hold promise in addressing complex control problems. However, the application of these approaches to biomedical systems is still in its early stages. This work employs dynamics-informed neural-network controllers as an alternative approach to control of medical digital twins. As a first use case, we focus on the control of agent-based models (ABMs), a versatile and increasingly common modelling platform in biomedicine. The effectiveness of the proposed neural-network control methods is illustrated and benchmarked against other methods with two widely used ABMs. To account for the inherent stochastic nature of the ABMs we aim to control, we quantify uncertainty in relevant model and control parameters.This article is part of the theme issue 'Uncertainty quantification for healthcare and biological systems (Part 1)'.

Emulating computer models with high-dimensional count output.

Computer models are used to study the real world, and often contain a large number of uncertain input parameters, produce a large number of outputs, may be expensive to run and need calibrating to real-world observations to be useful for decision-making. Emulators are often used as cheap surrogates for the expensive simulator, trained on a small number of simulations to provide predictions with uncertainty at unseen inputs. In epidemiological applications, for example compartmental or agent-based models for modelling the spread of infectious diseases, the output is usually spatially and temporally indexed, stochastic and consists of counts rather than continuous variables. Here, we consider emulating high-dimensional count output from a complex computer model using a Poisson lognormal PCA (PLNPCA) emulator. We apply the PLNPCA emulator to output fields from a COVID-19 model for England and Wales and compare this to fitting emulators to aggregations of the full output. We show that performance is generally comparable, while the PLNPCA emulator inherits desirable properties, including allowing the full output to be predicted while capturing correlations between outputs, providing high-dimensional samples of counts that are representative of the true model output.This article is part of the theme issue 'Uncertainty quantification for healthcare and biological systems (Part 1)'.

Economic inequality is a crucial determinant of observed patterns of environmental migration

Environmental migration occurs across spatial and temporal scales and is influenced by the dynamics of individual and collective human behavior, as well as environmental stresses. Agent-based modeling (ABM) has demonstrated potential for studying such complex processes, especially where individual decision-making is important. We developed an ABM to simulate environmental shocks, labor markets, and household livelihoods and used it to study environmental migration in rural Bangladesh. Using a pattern-oriented modeling approach, we found that an economic model of migration decisions can robustly reproduce observed patterns so long as economic inequality is properly represented. The results were sensitive to both the functional form and the parameters characterizing the distribution of land-ownership within a community. We conclude that community-level inequality has significant implications for environmental migration dynamics and should be considered in future policy interventions and research efforts.

Consumer attention and market concentration in e-commerce: an agent-based perspective

Abstract This study explores how limited consumer attention influences market concentration in e-commerce. Consumer attention, a scarce resource amidst abundant product information, plays a crucial role in shaping market dynamics. Despite its importance, the effect of limited consumer attention on e-commerce market concentration has not been extensively studied. Using agent-based modeling, we examine the interplay of consumer behaviors, bounded rationality, and social interactions in complex markets. Our results reveal that e-commerce market concentration persists even without product differentiation among sellers. Notably, a negative correlation emerges between consumer attention and market concentration, consistent across different market sizes and social connection densities. These findings provide theoretical insights into market concentration patterns in the Internet economy and contribute to the broader understanding of market structure dynamics.

Some problems of causal inference in agent-based macroeconomics

Abstract Guerini and Moneta (2017) have developed a sophisticated method of providing empirical evidence in support of the relations of causal dependence that macroeconomists engaging in agent-based modelling believe obtain in the target system of their models. The paper presents three problems that get in the way of successful applications of this method: problems that have to do with the potential chaos of the target system, the non-measurability of variables standing for individual or aggregate expectations, and the failure of macroeconomic aggregates to screen off individual expectations from the microeconomic quantities that constitute the aggregates. The paper also discusses the in-principle solvability of the three problems and uses a prominent agent-based model (the Keynes + Schumpeter model of the macroeconomy) as a running example.

Agent-Based Modeling in Economics and Finance: Past, Present, and Future

Agent-based modeling (ABM) is a novel computational methodology for representing the behavior of individuals in order to study social phenomena. Its use is rapidly growing in many fields. We review ABM in economics and finance and highlight how it can be used to relax conventional assumptions in standard economic models. ABM has enriched our understanding of markets, industrial organization, labor, macro, development, public policy, and environmental economics. In financial markets, substantial accomplishments include understanding clustered volatility, market impact, systemic risk, and housing markets. We present a vision for how ABMs might be used in the future to build more realistic models of the economy and review some of hurdles that must be overcome to achieve this. (JEL C63, D00, E00, G00)

Factors to scale out innovative organic farming systems: A case study in Flanders region, Belgium

Factors to scale out innovative organic farming systems: A case study in Flanders region, Belgium

Agent-based modelling: A stochastic approach to assessing personal exposure to environmental pollutants - Insights from the URBANOME project.

Agent-based modelling: A stochastic approach to assessing personal exposure to environmental pollutants - Insights from the URBANOME project.

Modeling residents’ long-term adaptation to geohazards in mountainous regions using agent-based models and Bayesian networks

Modeling residents’ long-term adaptation to geohazards in mountainous regions using agent-based models and Bayesian networks

Agent-based modeling in urban human mobility: A systematic review

Agent-based modeling in urban human mobility: A systematic review

Improving evacuation policies through agent-based modeling and stakeholder engagement in hazard-prone areas

Improving evacuation policies through agent-based modeling and stakeholder engagement in hazard-prone areas

Increasing child vaccination coverage can reduce influenza cases across age groups: An agent-based modeling study.

Increasing child vaccination coverage can reduce influenza cases across age groups: An agent-based modeling study.

Evaluating neighbourhood roads through agent-based modelling: A step towards the optimal pedestrian desire path system

As a part of road networks, sidewalks are critical in urban development and people’s lives. This study introduces a prototype system, the Road Evaluation by Desire Path Simulation System (RED-PaSS), based on agent-based modelling (ABM) designed to assess the alignment of road networks with pedestrian natural movement. RED-PaSS simulates pedestrian movement to generate an optimal desire path system, representing the most direct and natural paths that pedestrians would choose in urban settings. The model compares existing road networks against these simulated desire paths using key metrics, including road patch distances, segment alignment with desire paths, and origin–destination (OD) flow efficiency. This model enables urban planners to identify and address gaps in pedestrian infrastructure effectively. In the case study, RED-PaSS was applied to evaluate and rank 708 neighbourhood-scale road networks across the United States using high-resolution population density data from Meta and open-source road datasets from Zillow. The evaluation revealed that the roads closely aligned with simulated desire paths offer more direct and efficient pedestrian routes, enhancing natural walking movement. In contrast, poorly aligned sidewalks often force pedestrians to take indirect routes, reducing walkability and efficiency. In a sample neighbourhood, the evaluation showed that incorporating simulated desire paths reduced average pedestrian travel distances by approximately 28 %, from 43 m to 31 m, in sample neighborhoods. For the majority of the 708 evaluated road networks, the ranking index followed a Zipf’s law distribution with a value of R2=0.95, indicating a strong statistical correlation in the alignment of existing road networks with optimal desire paths. The top-ranked neighbourhoods demonstrated higher alignment scores and better coverage, effectively connecting areas with high pedestrian demand. By identifying discrepancies between existing road networks and optimal pedestrian desire paths, RED-PaSS provides actionable insights for designing more walkable urban environments and can potentially serve as a valuable tool for urban planners seeking to enhance urban road systems. Moreover, this work contributes to urban planning by offering a method to improve pedestrian network efficiency, aligning infrastructure with actual pedestrian natural movement behaviours, and emphasizing the need for integrating organic pedestrian movements into formal road network designs to create more functional and walkable cities.

Advancing epidemic modeling: The role of LLMs and generative agent-based models Comment on LLMs and generative agent-based models for complex systems research by Lu et al.

Advancing epidemic modeling: The role of LLMs and generative agent-based models Comment on LLMs and generative agent-based models for complex systems research by Lu et al.

On evolution of non-binding commitments: Comment on "LLMs and generative agent-based models for complex systems research" by Y. Lu et al.

On evolution of non-binding commitments: Comment on "LLMs and generative agent-based models for complex systems research" by Y. Lu et al.

A Retrospective Evaluation of Pandemic Policy Impact on University Campus: An Agent-based Modeling Approach for Mobility, Disease Propagation, and Testing During COVID-19

A Retrospective Evaluation of Pandemic Policy Impact on University Campus: An Agent-based Modeling Approach for Mobility, Disease Propagation, and Testing During COVID-19

Electrochemical immunosensor based on NiO for CA 15–3 detection and agent-based modelling for treatment optimization

Electrochemical immunosensor based on NiO for CA 15–3 detection and agent-based modelling for treatment optimization