A flexible bandwidth resource provisioning system with agent-enhanced SLA negotiation

Paper describes an effcicient approach for provisioning of network resources based on SLAs and a range of negotiating agents. The work arose from direct collboration with Fujitsu research and invlolved a worldwide press reslease of their agent brokering system which was based on this; also, a plenary address: A.Marshall (QUB) & A.Campbell (Columbia, USA) at 4th IFIP/IEEE International conference on Management of Multimedia Networks and Services' 2001 (MMNS'01). ISSN: 0926-6801

Computationally efficient sequential learning algorithms for direct link resource-allocating networks

Computionally efficient sequential learning algorithms are developed for direct-link resource-allocating networks (DRANs). These are achieved by decomposing existing recursive training algorithms on a layer by layer and neuron by neuron basis. This allows network weights to be updated in an efficient parallel manner and facilitates the implementation of minimal update extensions that yield a significant reduction in computation load per iteration compared to existing sequential learning methods employed in resource-allocation network (RAN) and minimal RAN (MRAN) approaches. The new algorithms, which also incorporate a pruning strategy to control network growth, are evaluated on three different system identification benchmark problems and shown to outperform existing methods both in terms of training error convergence and computational efficiency. (c) 2005 Elsevier B.V. All rights reserved.

Fighting for shells: how private information about resource value changes hermit crab pre-fight displays and escalated fight behaviour

Pre-fight displays typically provide honest, but sometimes dishonest, information about resource holding potential and may be influenced by assessment of resource value and hence motivation to acquire the resource. These assessments of potential costs and benefits are also predicted to influence escalated fight behaviour. This is examined in shell exchange contests of hermit crabs in which we establish an information asymmetry about a particularly poor quality shell. The poor shell was created by gluing sand to the interior whereas control shells lacked sand and the low value of the poor shell could not be accurately assessed by the opponent. Crabs in the poor shell showed changes in the use of pre-fight displays, apparently to increase the chances of swapping shells. When the fights escalated, crabs in poor shells fought harder if they took the role of attacker but gave up quickly if in the defender role. These tactics appear to be adaptive but do not result in a major shift in the roles taken or outcome. We thus link resource assessment with pre-fight displays, the roles taken, tactics used during escalation and the outcome of these contests.

Comparative analysis of dna methylation profiles in peripheral blood leukocytes versus lymphoblastoid cell lines

Previous reports have shown that DNA methylation profiles within primary human malignant tissues are altered when these cells are transformed into cancer cell lines. However, it is unclear if similar differences in DNA methylation profiles exist between DNA derived from peripheral blood leukocytes (PBLs) and corresponding Epstein-Barr Virus transformed lymphoblastoid cell lines (LCLs). To assess the utility of LCLs as a resource for methylation studies we have compared DNA methylation profiles in promoter and 5' regions of 318 genes in PBL and LCL sample pairs from patients with type 1 diabetes with or without nephropathy. We identified a total of 27 (similar to 8%) genes that revealed different DNA methylation profiles in PBL compared with LCL-derived DNA samples. In conclusion, although the profiles for most promoter regions were similar between PBL-LCL pairs, our results indicate that LCL-derived DNA may not be suitable for DNA methylation studies at least in diabetic nephropathy.

Analysis of the nearshore wave energy resource

Earlier studies have indicated that the gross nearshore wave energy resource is significantly smaller than the gross offshore wave energy resource implying that the deployment of wave energy converters in the nearshore is unlikely to be economic. However, it is argued that the gross wave energy resource is not an appropriate measure for determining the productivity of a wave farm and an alternative measure, the exploitable wave energy resource, is proposed. Calculation of a site's potential using the exploitable wave energy resource is considered superior because it accounts for the directional distribution of the incident waves and the wave energy plant rating that limits the power capture in highly energetic sea-states. A third-generation spectral wave model is used to model the wave transformation from deep water to a nearshore site in a water depth of 10 m. It is shown that energy losses result in a reduction of less than 10% of the net incident wave power. Annual wave data for the North Atlantic coast of Scotland is analysed and indicates that whilst the gross wave energy resource has reduced significantly by the 10 m depth contour, the exploitable wave energy resource is reduced by 7 and 22% for the two sites analysed. This limited reduction in exploitable wave energy resource means that for many exposed coasts, nearshore sites offer similar potential for exploitation of the wave energy resource as offshore sites.

Modelling Patient Duration of Stay to Facilitate Resource Management of Geriatric Hospitals

A fundamental aspect of health care management is the effective allocation of resources. This is of particular importance in geriatric hospitals where elderly patients tend to have more complex needs. Hospital managers would benefit immensely if they had advance knowledge of patient duration of stay in hospital. Managers could assess the costs of patient care and make allowances for these in their budget. In this paper, we tackle this important problem via a model which predicts the duration of stay distribution of patients in hospital. The approach uses phase-type distributions conditioned on a Bayesian belief network.