Agent-based models are used to analyse potential dynamics of complex (social) systems. Interactions of many heterogeneous actors and their environment are implemented on a computer, and the overall system evolution is observed in simulations in which these interactions are repeatedly carried out.
In an ABM, the behaviour of real-world actors (individuals but also other entities, e.g., firms) is represented by a set of computational rules that specify what an agent “does” in the simulations that represent the dynamical evolution of the system.
Computing is an activity that uses computers in order to manage and process information.
A domain-specific language (DSL) is a computer language specialized to a particular application domain. By limiting the focus to a particular domain, a “program” in a DSL can be understandable for domain experts, computer-checked, and at the same time possible to compile into efficient code (or other artefacts). Examples of DSLs include SQL for database queries, HTML for web pages, and MATLAB for matrix computations.
In an ABM, interactions by agents can take place in a common “world” around them; this environment may influence – and be influenced by – the agents. It may represent an ecological, geographical, institutional, etc. environment from the real world.
A software framework is an abstraction that provides generic functionality that can be applied or changed by specific application code. Consequently, a software framework provides a standardized frame to develop, deploy and execute applications.
Global Systems Science combines computer simulation modelling of complex social systems with stakeholder involvement to provide decision support for addressing global challenges, such as pandemics or climate change.
High Performance Computing is a specific niche within the Information and Communications Technology domain that can be used to understand extremely complex problem statements. It encompasses all activities that require extraordinary computational effort as well as highly challenging memory or storage capabilities.
any physical system that can be described in term of nodes (or vertices) and links (or edges) connecting them is a network.
Instances of networks can be found in various unrelated contexts: users on Facebook form a network establishing friendship relations with other users; the international trade from country to country can be described as a network; banks and financial institutions exchanging, for instance, loans can be analysed in terms of networks; a water supply is a network.
Depending on the nature of the interactions, networks can be undirected or directed (if the interaction from node A to node B is different from the one from B to A), weighted or binary (if the information about the intensity of the interaction is crucial), monopartite or multipartite (if all connections are allowed or if just those between different classes of nodes can exist).
In an ABM, there are a multitude of agents whose elementary behaviours depend on the behaviours of their neighbours. The interactions among agents can be described by a network.
In CoeGSS, three example global challenges were addressed as pilot studies: Health Habits, focusing on smoking habits and tobacco epidemics, Green Growth, analysing potential evolutions of the global car fleet and its emissions, and Global Urbanisation, studying the two-way relation between transport infrastructure decisions and price mechanisms, particularly concerning real-estate.
The CoEGSS Portal is the main entry point for the services provided by CoeGSS as a centre of excellence. It is a one-stop-shop where end users may find the tools and functionalities needed in their research of GSS such as yellow pages of people and entities involved in GSS research, matchmaking tool for finding the right stakeholders and services depending on your interests, data catalogue, training, discussion forums (in the way of Q&A) and simulations execution.
Scalability is the capability of a system, network, process or application to handle a growing amount of work. In High Performance Computing, two types of scaling have to be distinguished: strong scaling is defined as how the solution time varies with the number of processors for a fixed total problem size. In contrast, weak scaling is defined as how the solution time varies with the number of processors for a fixed problem size per processor. So scalability itself has two different scales, large-scale environments but also the individual building blocks of them.
Synthetic Population (ABM)
A synthetic population is a set of virtual agents, that reproduces statistical distributions of the real-world population under study for relevant features.