Chris Scogings

Structural circuits and attractors in Kauffman networks

There has been some ambiguity about the growth of attractors in Kauffman networks with network size. Some recent work has linked this to the role and growth of circuits or loops of boolean variables. Using numerical methods we have investigated the growth of structural circuits in Kauffman networks and suggest that the exponential growth in the number of structural circuits places a lower bound on the complexity of the growth of boolean dependency loops and hence of the number of attractors. We use a fast and exact circuit enumeration method that does not rely on sampling trajectories. We also explore the role of structural self-edges, or self-inputs in the NK-model, and how they affect the number of structural circuits and hence of attractors.

A method for the early stages of interactive system design using UML and Lean Cuisine

In interactive system design, models and notations are required for describing user tasks, and for describing the structure of the human-computer dialogue to support these tasks. These descriptions should ideally be linked. This paper examines task modelling in UML and dialogue description in Lean Cuisine+, and describes a method for the early stages of interactive systems design which incorporates both notations. This provides a means of representing tasks in the context of the structure of the user interface, i.e. of explicitly showing the transformation of tasks to dialogue.

Grid-boxing for spatial simulation performance optimisation

Computer simulations of complex systems such as physical aggregation processes or swarming and collective behaviour of life-forms, often require order N-squared computational complexity for N microscopic components. This is a significant handicap to simulating systems large enough to compare with real-world experimental data. We discuss space partitioning methods for two such simulation codes and demonstrate complexity improvements by taking advantage of information about locations and interaction distances of the microscopic model components. We present results for a diffusion limited cluster-cluster aggregation code and for an artificial life simulation code. We discuss the data structures necessary to support such algorithms and show how they can be implemented to obtain high performance and maximal simulation productivity for a given computational resource. There are some subtleties in whether such spatial partitioning algorithms should produce a computational complexity of N to some power between 1 and 2 or whether they should be order N log N. We discuss these effects in the context of our data.

An agent-based model of the Battle of Isandlwana

Agent-based models have been used to capture and analyze the essential behaviors of combat units although the number of agents used has been fairly low. We experiment with a microscopically detailed agent model in which over 20,000 soldiers are represented individually (one agent per soldier) in a simulation of the Battle of Isandlwana in 1879. We describe how a rule based model can be specified for soldiers on both sides and how it can be specialized for different skill sets and fighting capabilities of soldier agents belonging to particular units. We address some of the challenges of programming a model consisting of large numbers of agents. We demonstrate that our model provides a simulation of the battle with considerable historical accuracy and then go on to show how the same model can be used to demonstrate a plausible alternative to history.

Spatial pattern growth and emergent animat segregation

Spatial agent models can be used to explore self-organizing effects such as pattern growth and segregation. An approximate time line of key animat ideas and agent systems is presented and discussed. These ideas have led to a unique animat simulation model for studying emergence effects in artificial life systems and this predator-prey model is employed to study emergent behaviours in systems of up to around one million individual animat agents. The patterns, structures and emergent properties of the model are compared with the spatial patterns formed in non-intelligence based models that are governed only by statistical mechanics. An emergent species separation effect is found amongst the prey animats when a simple genetic marker is employed to track animats and introduce a microscopic breeding preference. Results are presented using quantitative metrics such as the animal spatial density and the pair-wise density-density correlation function. Ways in which these metrics can be used to categorize different self-organizational model regimes are discussed.

Spatial emergence of genotypical tribes in an animat simulation model

We observe the spontaneous emergence of spatial tribes in an animat agent model where simple genetic inheritance is supported. Our predator-prey model simulates a flat-world of animat agents which breed, move, eat and predate according to priorities encoded in their genotype. Initialising a random mixture of all possible priority list genotypes, we find not only that only a small fraction of possible genotypes are favoured for survival, but that distinct spatial patterns of different tribes emerge. We report on the emergent macroscopic features in our model and discuss their correspondent mapping to microscopic animat rules and genotypes. Even a simple gene-reordering mechanism gives rise to complex emergent behaviour.

User-friendly scheduling tools for large-scale simulation experiments

Planning and steering numerical experiments that involve many simulations are difficult tasks to automate. We describe how a simulation scheduling tool can help experimenters submit and revoke simulation jobs on the basis of the most up to date partial results and resource estimates. We show how ideas such as pre- and post-conditions; interrupt handling; rapid experiment schema creation; and sparse parameter cross-products can be used to make a generalisable and user- friendly scheduling toolset. We describe our prototype in the context of typical long-running computational experiments of a complex networks simulation problem.

Emergent Spatial Agent Segregation

Animat agents are usually formulated as spatially located agents that interact according to some microscopic behavioural rules. We use our predator-prey animat model to explore spatial segregation and other self-organising effects. We compare the emergent macroscopic behaviour with that of non-intelligence models such as those governed solely by microscopic statistical mechanics rules. We report on an emergent separation of sub-species amongst our prey animals when a very simple genetic marker is used and a microscopic breeding preference is introduced. We discuss some quantitative metrics such as the spatial density of animals and the density-density correlation function and how these can be used to categorize the different self-organisational regimes that emerge from the model.

Linking tasks, dialogue and GUI design: a method involving UML and Lean Cuisine+

Abstract Unified Modelling Language (UML) is rapidly becoming the international standard for software design notation and methods. However, it offers little or no support for the design of the user interface. Models and notations are required for describing user tasks and the structure of the human–computer dialogue to support these tasks. Lean Cuisine+ provides a notation for both dialogue and task modelling. This paper describes a method for the initial steps of interface design which incorporates both the UML and Lean Cuisine+ notations and provides a means of representing tasks in the context of the dialogue structure of the user interface.

Modelling the mock-up: towards the automatic specification of the behaviour of early prototypes

A shortcoming of current early prototyping tools is that although they permit the designer to construct a mock-up of the look and feel of the interface, they provide no model of the interaction. Such tools could be extended to produce such a model as a byproduct of the construction of the prototype. This research explores such an extension to HyperCard, and reviews its utility. The interaction model is specified in Lean Cuisine+, a semi-formal graphical notation for describing the behaviour of direct manipulation graphical user interfaces.