The purpose of the model is to study the dynamical relationship between individual needs and group performance when focusing on self-organizing task allocation. For this, we develop a model that formalizes Deci & Ryan’s self-determination theory (SDT) theory into an ABM creating a framework to study the social dynamics that pertain to the mutual relations between the individual and group level of team performance. Specifically, it aims to answer how the three individual motivations of autonomy, competence, and belonging affect team performance.

This model system aims to simulate the whole process of task allocation, task execution and evaluation in the team system through a feasible method. On the basis of Complex Adaptive Systems (CAS) theory and Agent-based Modelling (ABM) technologies and tools, this simulation system attempts to abstract real-world teams into MAS models. The author designs various task allocation strategies according to different perspectives, and the interaction among members is concerned during the task-performing process. Additionally, knowledge can be acquired by such an interaction process if members encounter tasks they cannot handle directly. An artificial computational team is constructed through ABM in this simulation system, to replace real teams and carry out computational experiments. In all, this model system has great potential for studying team dynamics, and model explorers are encouraged to expand on this to develop richer models for research.

This model aims to replicate the evolution of opinions and behaviours on a communal plan over time. It also aims to foster community dialogue on simulation outcomes, promoting inclusivity and engagement. Individuals (referred to as agents), grouped based on Sinus Milieus (Groh-Samberg et al., 2023), face a binary choice: support or oppose the plan. Motivated by experiential, social, and value needs (Antosz et al., 2019), their decision is influenced by how well the plan aligns with these fundamental needs.

The aim of our model is to investigate the team dynamics through two types of task allocation strategies, with a focus on the dynamic interplay between individual needs and group performance. To achieve this goal, we have formulated an agent-based model (ABM) to formalize Deci & Ryan’s self-determination theory (SDT) and explore the social dynamics that govern the relationship between individual and group levels of team performance.

This model visualizes gradient descent optimization - the fundamental algorithm used to train neural networks and other machine learning models. Agents represent different optimization algorithms searching for the minimum of a loss landscape (the “error surface” that ML models try to minimize during training).

The model demonstrates how different optimizer types (SGD, Momentum with different parameters) behave on various loss landscapes, from simple bowls to the notoriously difficult Rosenbrock “banana valley” function. This helps build intuition about why certain optimization algorithms work better than others for different problem geometries.

HOW IT WORKS

…

The model is an agent-based artificial stock market where investors connect in a dynamic network. The network is dynamic in the sense that the investors, at specified intervals, decide whether to keep their current adviser (those investors they receive trading advise from). The investors also gain information from a private source and share public information about the risky asset. Investors have different tendencies to follow the different information sources, consider differing amounts of history, and have different thresholds for investing.

This model was created to investigate the potential impacts of large-scale recreational and transport-related physical activity promotion strategies on six United Nations Sustainable Development Goals (SDGs) related outcomes—road traffic deaths (SDG 3), transportation mode share (SDG 9), convenient access to public transport, levels of fine particulate matter, and access to public open spaces (SDG 11), and levels of carbon dioxide emissions (SDG 13)—in three cities designed as abstract representations of common city types in high-, middle-, and low-income countries.

The SIM-VOLATILE model is a technology adoption model at the population level. The technology, in this model, is called Volatile Fatty Acid Platform (VFAP) and it is in the frame of the circular economy. The technology is considered an emerging technology and it is in the optimization phase. Through the adoption of VFAP, waste-treatment plants will be able to convert organic waste into high-end products rather than focusing on the production of biogas. Moreover, there are three adoption/investment scenarios as the technology enables the production of polyhydroxyalkanoates (PHA), single-cell oils (SCO), and polyunsaturated fatty acids (PUFA). However, due to differences in the processing related to the products, waste-treatment plants need to choose one adoption scenario.

In this simulation, there are several parameters and variables. Agents are heterogeneous waste-treatment plants that face the problem of circular economy technology adoption. Since the technology is emerging, the adoption decision is associated with high risks. In this regard, first, agents evaluate the economic feasibility of the emerging technology for each product (investment scenarios). Second, they will check on the trend of adoption in their social environment (i.e. local pressure for each scenario). Third, they combine these two economic and social assessments with an environmental assessment which is their environmental decision-value (i.e. their status on green technology). This combination gives the agent an overall adaptability fitness value (detailed for each scenario). If this value is above a certain threshold, agents may decide to adopt the emerging technology, which is ultimately depending on their predominant adoption probabilities and market gaps.

A dynamic identity model for misinformation in social networks, an agent-based model of social identity and misinformation dynamics.

I developed this model as a part of my master’s thesis, “Does social identity drive belief and persistence in online misinformation? An agent-based modelling approach” at University College Dublin, Ireland (2024-2025).

The purpose of this model is to further understand the dynamics of misinformation sharing as an expression of social identity. I introduce a framework to understand the influence of self-categorisation on misinformation persistence in social network. It integrates a social learning model with the Dynamic Identity Model for Agents (DIMA) using simple logic to simulate the social trade-offs driving misinformation and observe the effects on misinformation spread.

This is a modification of a model published previous by Barton and Riel-Salvatore (2012). In this model, we simulate six regional populations within Last Glacial Maximum western Europe. Agents interact through reproduction and genetic markers attached to each of six regions mix through subsequent generations as a way to track population dynamics, mobility, and gene flow. In addition, the landscape is heterogeneous and affects agent mobility and, under certain scenarios, their odds of survival.