Intelligent agents for fault tolerance: from multi-agent simulation to cluster-based implementation

Recent research in multi-agent systems incorporate fault tolerance concepts, but does not explore the extension and implementation of such ideas for large scale parallel computing systems. The work reported in this paper investigates a swarm array computing approach, namely 'Intelligent Agents'. A task to be executed on a parallel computing system is decomposed to sub-tasks and mapped onto agents that traverse an abstracted hardware layer. The agents intercommunicate across processors to share information during the event of a predicted core/processor failure and for successfully completing the task. The feasibility of the approach is validated by simulations on an FPGA using a multi-agent simulator, and implementation of a parallel reduction algorithm on a computer cluster using the Message Passing Interface.

Queue-based agent architecture for multimodal interfaces

This paper presents a queue-based agent architecture for multimodal interfaces. Using a novel approach to intelligently organise both agents and input data, this system has the potential to outperform current state-of-the-art multimodal systems, while at the same time allowing greater levels of interaction and flexibility. This assertion is supported by simulation test results showing that significant improvements can be obtained over normal sequential agent scheduling architectures. For real usage, this translates into faster, more comprehensive systems, without the limited application domain that restricts current implementations.

Data based constructive identification - overcoming the curse of dimensionality

The modelling of a nonlinear stochastic dynamical processes from data involves solving the problems of data gathering, preprocessing, model architecture selection, learning or adaptation, parametric evaluation and model validation. For a given model architecture such as associative memory networks, a common problem in non-linear modelling is the problem of "the curse of dimensionality". A series of complementary data based constructive identification schemes, mainly based on but not limited to an operating point dependent fuzzy models, are introduced in this paper with the aim to overcome the curse of dimensionality. These include (i) a mixture of experts algorithm based on a forward constrained regression algorithm; (ii) an inherent parsimonious delaunay input space partition based piecewise local lineal modelling concept; (iii) a neurofuzzy model constructive approach based on forward orthogonal least squares and optimal experimental design and finally (iv) the neurofuzzy model construction algorithm based on basis functions that are Bézier Bernstein polynomial functions and the additive decomposition. Illustrative examples demonstrate their applicability, showing that the final major hurdle in data based modelling has almost been removed.

Queue-based agent communication architecture for intelligent human-like computer interfaces

Human-like computer interaction systems requires far more than just simple speech input/output. Such a system should communicate with the user verbally, using a conversational style language. It should be aware of its surroundings and use this context for any decisions it makes. As a synthetic character, it should have a computer generated human-like appearance. This, in turn, should be used to convey emotions, expressions and gestures. Finally, and perhaps most important of all, the system should interact with the user in real time, in a fluent and believable manner.

On a stochastic Trotter formula with application to spontaneous localization models

We consider the relation between so called continuous localization models—i.e. non-linear stochastic Schrödinger evolutions—and the discrete GRW-model of wave function collapse. The former can be understood as scaling limit of the GRW process. The proof relies on a stochastic Trotter formula, which is of interest in its own right. Our Trotter formula also allows to complement results on existence theory of stochastic Schrödinger evolutions by Holevo and Mora/Rebolledo.

Risk assessment of UK skylark populations using life-history and individual-based landscape models

Abstract: Following a workshop exercise, two models, an individual-based landscape model (IBLM) and a non-spatial life-history model were used to assess the impact of a fictitious insecticide on populations of skylarks in the UK. The chosen population endpoints were abundance, population growth rate, and the chances of population persistence. Both models used the same life-history descriptors and toxicity profiles as the basis for their parameter inputs. The models differed in that exposure was a pre-determined parameter in the life-history model, but an emergent property of the IBLM, and the IBLM required a landscape structure as an input. The model outputs were qualitatively similar between the two models. Under conditions dominated by winter wheat, both models predicted a population decline that was worsened by the use of the insecticide. Under broader habitat conditions, population declines were only predicted for the scenarios where the insecticide was added. Inputs to the models are very different, with the IBLM requiring a large volume of data in order to achieve the flexibility of being able to integrate a range of environmental and behavioural factors. The life-history model has very few explicit data inputs, but some of these relied on extensive prior modelling needing additional data as described in Roelofs et al.(2005, this volume). Both models have strengths and weaknesses; hence the ideal approach is that of combining the use of both simple and comprehensive modeling tools.

Association between serotonin 5-HT-2C receptor gene (HTR2C) polymorphisms and obesity- and mental health-related phenotypes in a large population-based cohort

OBJECTIVE: Studies have shown that common single-nucleotide polymorphisms (SNPs) in the serotonin 5-HT-2C receptor (HTR2C) are associated with antipsychotic agent-induced weight gain and the development of behavioural and psychological symptoms. We aimed to analyse whether variation in the HTR2C is associated with obesity- and mental health-related phenotypes in a large population-based cohort. METHOD: Six tagSNPs, which capture all common genetic variation in the HTR2C gene, were genotyped in 4978 men and women from the European Prospective Investigation into Cancer (EPIC)-Norfolk study, an ongoing prospective population-based cohort study in the United Kingdom. To confirm borderline significant associations, the -759C/T SNP (rs3813929) was genotyped in the remaining 16 003 individuals from the EPIC-Norfolk study. We assessed social and psychological circumstances using the Health and Life Experiences Questionnaire. Genmod models were used to test associations between the SNPs and the outcomes. Logistic regression was performed to test for association of SNPs with obesity- and mental health- related phenotypes. RESULTS: Of the six HTR2C SNPs, only the T allele of the -759C/T SNP showed borderline significant associations with higher body mass index (BMI) (0.23 kg m(-2); (95% confidence interval (CI): 0.01-0.44); P=0.051) and increased risk of lifetime major depressive disorder (MDD) (Odds ratio (OR): 1.13 (95% CI: 1.01-1.22), P=0.02). The associations between the -759C/T and BMI and lifetime MDD were independent. As associations only achieved borderline significance, we aimed to validate our findings on the -759C/T SNP in the full EPIC-Norfolk cohort (n=20 981). Although the association with BMI remained borderline significant (beta=0.20 kg m(-2); 95% CI: 0.04-0.44, P=0.09), that with lifetime MDD (OR: 1.01; 95% CI: 0.94-1.09, P=0.73) was not replicated. CONCLUSIONS: Our findings suggest that common HTR2C gene variants are unlikely to have a major role in obesity- and mental health-related traits in the general population.

Evaluation of the Plant–Craig stochastic convection scheme (v2.0) in the ensemble forecasting system MOGREPS-R (24 km) based on the Unified Model (v7.3)

The Plant–Craig stochastic convection parameterization (version 2.0) is implemented in the Met Office Regional Ensemble Prediction System (MOGREPS-R) and is assessed in comparison with the standard convection scheme with a simple stochastic scheme only, from random parameter variation. A set of 34 ensemble forecasts, each with 24 members, is considered, over the month of July 2009. Deterministic and probabilistic measures of the precipitation forecasts are assessed. The Plant–Craig parameterization is found to improve probabilistic forecast measures, particularly the results for lower precipitation thresholds. The impact on deterministic forecasts at the grid scale is neutral, although the Plant–Craig scheme does deliver improvements when forecasts are made over larger areas. The improvements found are greater in conditions of relatively weak synoptic forcing, for which convective precipitation is likely to be less predictable.