Search

Machine learning (ML) approach to modeling\ud and predicting real-world dynamic system behaviours has\ud received widespread research interest. While ML capability in\ud approximating any nonlinear or complex system is promising,\ud it is often a black-box approach, which lacks the physical\ud meanings of the actual system structure and its parameters, as\ud well as their impacts on the system. This paper establishes a\ud model to provide explanation on how system parameters affect\ud its output(s), as such knowledge would lead to potential useful,\ud interesting and novel information. The paper builds on our\ud previous work in machine learning, and also combines an\ud evolutionary artificial neural networks with sensitivity analysis\ud to extract and validate key factors affecting the cloud data\ud center energy performance. This provides an opportunity for\ud software analyst to design and develop energy-aware\ud applications and for Hadoop administrator to optimize the\ud Hadoop infrastructure by having Big Data partitioned in\ud bigger chunks and shortening the time to complete MapReduce\ud jobs.

The success of Hadoop, an open-source\ud framework for massively parallel and distributed computing, is\ud expected to drive energy consumption of cloud data centers to\ud new highs as service providers continue to add new\ud infrastructure, services and capabilities to meet the market\ud demands. While current research on data center airflow\ud management, HVAC (Heating, Ventilation and Air\ud Conditioning) system design, workload distribution and\ud optimization, and energy efficient computing hardware and\ud software are all contributing to improved energy efficiency,\ud energy forecast in cloud computing remains a challenge. This\ud paper reports an evolutionary computation based modeling\ud and forecasting approach to this problem. In particular, an\ud evolutionary neural network is developed and structurally\ud optimized to forecast the energy load of a cloud data center.\ud The results, both in terms of forecasting speed and accuracy,\ud suggest that the evolutionary neural network approach to\ud energy consumption forecasting for cloud computing is highly\ud promising.

Sleep deprivation (SD) can give rise to faltering attention but the mechanics underlying this remain uncertain. Using a covert attention task that required attention to a peripheral target location, we compared the effects of attention and SD on baseline activity prior to visual stimulation as well as on stimulus-evoked activity. Volunteers were studied after a night of normal sleep (RW) and a night of SD. Baseline signal elevations evoked by preparatory attention in the absence of visual stimulation were attenuated within rFEF, rIPS (sparing SEF) and all retinotopically mapped visual areas during SD, indicative of impaired endogenous attention. In response to visual stimuli, attention modulated activation in higher cortical areas and extrastriate cortex (hV4, ventral occipital areas) after RW. SD attenuated rFEF, rIPS, V3a and VO stimulus-evoked activation regardless of whether stimuli were attended. Notably, the modulation of stimulus-evoked activation by attention was not affected by SD unlike for the preparatory period, suggesting a reduced number, but still functional circuits during SD. Deficits in endogenous attention in SD dominate in the preparatory period, whereas changes in stimulus-related activation arise from an interaction between compromised fronto-parietal top-down control of attention and reduced sensitivity of extrastriate visual cortex to top-down or bottom-up inputs.

'Eco-friendly shipping’ and fuel efficiency are important issues and challenges faced by marine and offshore industry today. This is primary due to increasing stringent environmental regulations and the pressure to reduce fuel cost to stay competitive. Furthermore, the technical requirements of offshore vessels have also increased significantly. Recent developments in advance optimisation techniques and High Performance Computing (HPC) are key enablers for Simulation Based Design (SBD) in ship and offshore applications, where overall efficiency of design process and hydrodynamic performance can be greatly improved. This paper integrates a geometry modification technique with advance optimisation and evaluation algorithms, with the aim to improve the overall efficiency and hydrodynamic performance of ‘ship-shaped’ offshore vessels. By incorporating concepts from Free-Form Deformation (FFD) and Genetic Algorithm (GA), the proposed approach is applied to optimise the hull form design of a heavy-lift and pipe-lay vessel to reduce the overall resistance while improving its sea-keeping performance.

Multi-step micromechanics-based models are developed to predict the overall effective elastic moduli of porous ceramic with randomly oriented carbon nanotube (CNT) reinforcements. The presence of porosity in the ceramic matrix that has been previously neglected in the literature is considered in present analysis. The ceramic matrix with porosity is first homogenized using a classical Mori-Tanaka model. Then, the homogenized porous ceramic matrix with randomly oriented CNTs is analysed using two micromechanics models. The results predicted by the present models are compared with experimental and analytical results that have been reported in literature. The comparison shows that the discrepancies between the present analytical results and experimental data are about 10% for 4 wt% of CNTs and about 0.5% for 8 wt% CNTs, both substantially lower than the discrepancies currently reported in the literature.

New environmental regulations and volatile fuel\ud prices have resulted in an ever-increasing need for reduction\ud in carbon emission and fuel consumption. Designs of marine\ud and offshore vessels are more demanding with complex\ud operating requirements and oil and gas exploration\ud venturing into deeper waters and hasher environments.\ud Combinations of these factors have led to the need to\ud optimise the design of the hull for the marine and offshore\ud industry. The contribution of this paper is threefold. Firstly,\ud the paper provides a comprehensive review of the state-ofthe-\ud art techniques in hull form design. Specifically, it\ud analyses geometry modelling, shape transformation,\ud optimisation and performance evaluation. Strengths and\ud weaknesses of existing solutions are also discussed.\ud Secondly, key challenges of hull form optimisation specific\ud to the design of marine and offshore vessels are identified\ud and analysed. Thirdly, future trends in performing hull\ud form design optimisation are investigated and possible\ud solutions proposed. A case study on the design optimisation\ud of bulbous bow for passenger ferry vessel to reduce wavemaking\ud resistance is presented using NAPA software.\ud Lastly, main issues and challenges are discussed to stimulate\ud further ideas on future developments in this area, including\ud the use of parallel computing and machine intelligence.

Recent studies have provided coexistence and interaction models between IEEE 802.15.4 and IEEE 802.11 standards. However, the performance of IEEE 802.15.4 devices under WiFi interference are evaluated based on limit parameters i.e. Packet Reception Rate, which does not exhibit the dynamic interactions in the wireless channel. In this paper, we conduct a series of experiments to demonstrate the dynamic interactions between the IEEE 802.15.4 and IEEE 802.11 bgn standards on relevant devices. The performance of four existing Link Quality Estimators (LQEs) of IEEE 802.15.4 nodes under the IEEE 802.11 bgn interference is analyzed. We show that IEEE 802.15.4 transmission failures are largely due to channel access failures rather than corrupted data packets. Based on the analysis, we propose a new LQE - Packet Reception Rate with Clear Channel Assessment - by merging the Clear Channel Assessment count with the Packet Reception Rate. In comparison to existing LQEs, results show that the new estimator distinguishes persistent IEEE 802.11 bgn traffic more robustly.

Amongst renewable generators, photovoltaics (PV) are becoming more popular as the appropriate low cost solution to meet increasing energy demands. However, the integration of renewable energy sources to the electricity grid possesses many challenges. The intermittency of these non-conventional sources often requires accurate forecast, planning and optimal management. Many attempts have been made to tackle these challenges; nonetheless, existing methods fail to accurately capture the underlying characteristics of the system. There exists scope to improve present PV yield forecasting models and methods. This paper explores the use of apriori knowledge of PV systems to build clear box models and identify uncertain parameters via heuristic algorithms. The model is further enhanced by incorporating black box models to account for unmodeled uncertainties in a novel grey-box forecasting and modeling of PV systems.

Among renewable generators, photovoltaics (PV) is showing an increasing suitability and a lowering cost. However, integration of renewable energy sources possesses many challenges, as the intermittency of these non-conventional sources often requires generation forecast, planning and optimal management. There exists scope to improve present PV yield forecasting models and methods. For example, the popular dynamic neural network modelling method suffers from the lack of a selection mechanism for an optimal network structure. This paper develops an enhanced network for short-term forecasting of PV power yield, termed a `focused time-delay neural network' (FTDNN). The problem of optimizing the FTDNN structure is reduced to optimizing the number of delay steps and the number of neurons in the hidden layer alone and this problem is conveniently solved through heuristics. Two such algorithms, a genetic algorithm and particle swarm optimization (PSO) have been tested and both prove efficient and can improve the forecasting accuracy of the dynamic network. Given the success of the PSO in solving this discontinuous structural optimization problem, it is expected that PSO offers potential in optimizing both the structure and parameters of a forecasting model.

The importance of predicting the output power of Photovoltaic (PV) plants is crucial in modern power system applications. Predicting the power yield of a PV generation system helps the process of dispatching the power into a grid with improved efficiency in generation planning and operation. This work proposes the use of intelligent tools to forecast the real power output of PV units. These tools primarily comprise dynamic neural networks which are capable of time-series predictions with good reliability. This paper begins with a brief review of various methods of forecasting solar power reported in literature. Results of preliminary work on a 5kW PV panel at King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia, is presented. Focused Time Delay and Distributed Time Delay Neural Networks were used as a forecasting tool for this study and their performance was compared with each other.

Approximately 10% of women with invasive epithelial ovarian cancer (EOC) carry deleterious germline mutations in BRCA1 or BRCA2. A recent article suggested that BRCA2-related EOC was associated with an improved prognosis, but the effect of BRCA1 remains unclear.To characterize the survival of BRCA carriers with EOC compared with noncarriers and to determine whether BRCA1 and BRCA2 carriers show similar survival patterns.A pooled analysis of 26 observational studies on the survival of women with ovarian cancer, which included data from 1213 EOC cases with pathogenic germline mutations in BRCA1 (n = 909) or BRCA2 (n = 304) and from 2666 noncarriers recruited and followed up at variable times between 1987 and 2010 (the median year of diagnosis was 1998).Five-year overall mortality.The 5-year overall survival was 36% (95% CI, 34%-38%) for noncarriers, 44% (95% CI, 40%-48%) for BRCA1 carriers, and 52% (95% CI, 46%-58%) for BRCA2 carriers. After adjusting for study and year of diagnosis, BRCA1 and BRCA2 mutation carriers showed a more favorable survival than noncarriers (for BRCA1: hazard ratio [HR], 0.78; 95% CI, 0.68-0.89; P.001; and for BRCA2: HR, 0.61; 95% CI, 0.50-0.76; P.001). These survival differences remained after additional adjustment for stage, grade, histology, and age at diagnosis (for BRCA1: HR, 0.73; 95% CI, 0.64-0.84; P.001; and for BRCA2: HR, 0.49; 95% CI, 0.39-0.61; P.001). The BRCA1 HR estimate was significantly different from the HR estimated in the adjusted model (P for heterogeneity = .003).Among patients with invasive EOC, having a germline mutation in BRCA1 or BRCA2 was associated with improved 5-year overall survival. BRCA2 carriers had the best prognosis.

Introduction: The stress -vulnerability model (Zubin et al. 1977) is and extremely useful model for identifying and treating relapses of mental illness. We accept that human persons carry genetic and other predisposition to mental illness. However, the question arises ast to how stress impacts pn a person in order to cause mentsl illness to develop. Furthermore there arises the issue as to what other effects such stress has on the human body beyond the human brain. Our aim was to research and integrate the current literature in order to establish how stress impacts on the brain at the cellular level, and to establish wheather there are other consequences for the human body brought about by the impact of stress on the human brain. Method: Literature Search, using pubmed. Results: We have identified much literature on how stress affects biological mechanisms within the brain, and how it relates to biological vulnerabilities carried by different individuals. Conclusion: We have identified communalities in how the interplay between stress and vulnerability occurs in different disease processes.

The number of people that now go on to develop Alzheimer's disease (AD) and other types of dementia is rapidly rising. For maximum benefits from new treatments, the disease should be diagnosed as early as possible, but this is difficult with current clinical criteria. Potentially, the EEG can serve as an objective, first line of decision support tool to improve diagnosis. It is non-invasive, widely available, low-cost and could be carried out rapidly in the high-risk age group that will develop AD. Changes in the EEG due to the dementing process could be quantified as an index or marker. In this paper, we investigate two information theoretic methods (Tsallis entropy and universal compression algorithm) as a way to generate potentially robust markers from the EEG. The hypothesis is that the information theoretic makers for AD are significantly different to those of normal subjects. An attraction of the information theoretic approach is that, unlike most existing methods, there may be a natural link between the underlying ideas of information theoretic methods, the physiology of AD and its impact on brain functions. Data compression has not been investigated as a means of generating EEG markers before and is attractive because it does not require a priori knowledge of the source model. In this paper, we focus on the LZW algorithm because of its sound theoretical foundation. We used the LZW algorithm and Tsallis model to compute the markers (compression ratios and normalized entropies, respectively) from two EEG datasets. The results show that the information theoretic methods can be used to compute EEG markers for AD.

Worldwide, the number of people that develop Alzheimer's disease and other types of dementia is rapidly rising and will create a considerable financial burden on the health and social services. The availability of new drugs that may slow or even halt the disease progression makes accurate early detection crucial. Objective methods are needed to support clinical diagnosis and care for patients; to quantify severity, monitor progression and response to new treatments. Electrophysiological markers have an important role to play in the objective assessment and care for dementia. The EEG provides a measure of brain dysfunction and EEG changes could be detected fairly early in the dementing process. Subject-specific EEG analysis offers the possibility of using objective methods to assess and care for dementia on an individual basis. The main objectives of this paper are: (i) to introduce the concepts of subject-specific EEG analysis as a basis for improving diagnosis and care for dementia; and (ii) present two novel methods for deriving suitable subject-specific electrophysiological markers - analysis of fractal dimension and zero crossing interval density of the EEG. We present findings that indicate that the methods are potentially good candidates for the development of individualized, low-cost, easy to administer and reasonably accurate methods for detecting dementia within the growing at risk population

This paper makes an outline case for the need for a low-cost, easy to administer method for detecting dementia within the growing at risk population. It proposes two methods for electroencephalogram (EEG) analysis for detecting dementia that could fulfil such a need. The paper describes a fractal dimension-based method for analyzing the EEG waveforms of subjects with dementia and reports on an assessment which demonstrates that an appropriate fractal dimension measure could achieve 67% sensitivity to probable Alzheimer's disease (as suggested by clinical psychometric testing and EEG findings) with a specificity of 99.9%. An alternative method based on the probability density function of the zero-crossing intervals is shown to achieve 78% sensitivity to probable Alzheimer's disease and an estimated sensitivity to probable Vascular (or mixed) dementia of 35% (as suggested by clinical psychometric testing and EEG findings) with a specificity of 99.9%. This compares well with other studies, reported by the American Academy of Neurology, which typically provide a sensitivity of 81% and specificity of 70%. The EEG recordings used to assess these methods included artefacts and had no a priori selection of elements "suitable for analysis." This approach gives a good prediction of the usefulness of the methods, as they would be used in practice. A total of 39 patients (30 probable Alzheimer's Disease, six Vascular Dementia and three mixed dementia) and 42 healthy volunteers were involved in the study. However, although results from the preliminary evaluation of the methods are promising, there is a need for a more extensive study to validate the methods using EEGs from a larger and more varied patient cohorts with neuroimaging results, to exclude other causes and cognitive scores to correlate results with severity of cognitive status.

Summarization: The primary aim of this paper is to present a new concept, bioprofiling over Grid, and to illustrate how Grid computing may be used to support individualisation of healthcare in future, with the aid of a new test bed, the BIOPATTERN Grid. The BIOPATTERN Grid is designed to facilitate seamless sharing of geographically distributed bioprofile databases and to support the analysis of bioprofiles to combat major diseases such as brain diseases and cancer within a major EU project, BIOPATTERN (www.biopattern.org). The main objectives in this paper are 1) to report the development of the BIOPATTERN Grid for biopattern analysis and bioprofiling in support of individualisation of healthcare; 2) to illustrate how the BIOPATTERN Grid could be used for biopattern analysis and bioprofiling for early detection of dementia. We present the architecture and general functionalities of BIOPATTERN Grid, and the development of a prototype test bed (including a Grid Portal and Grid services for early detection of dementia). We illustrate the concept of bioprofiling over Grid and discuss issues such as scalability in high through-put computing for biodata analysis associated with bioprofiling for dementia. Results show benefits in both high throughput computing in biodata analysis and for individualisation of healthcare using Grid computing which makes it possible to access geographically distributed patient's information for subject-specific data analysis for early detection of dementia. Παρουσιάστηκε στο: 1st international conference on Scalable information systems

Rare germline mutations in the breast and ovarian cancer predisposition genes BRCA1 and BRCA2 are present in roughly 5 percent of ovarian cancer patients. Both genes play key roles in DNA damage repair but appear to have distinct, although often complementary, functions. The risks of breast and ovarian cancer have been shown to differ between BRCA1 and BRCA2 carriers, and mutation-specific effects have also been suggested for both groups. Published reports comparing the survival of ovarian cancer patients with and without BRCA1/2 mutations generally show a survival advantage among mutation carriers. This is thought to be related to an improved response of carriers to platinum-based therapies. Since it is not known whether BRCA1 and BRCA2 carriers show similar survival patterns, we have performed a large, multicenter study to investigate the impact of germline BRCA1 and BRCA2 mutations on ovarian cancer survival. 3,531 invasive epithelial ovarian cancer cases (1,178 BRCA1, 367 BRCA2 and 1,986 BRCA-negative) with survival time data from twenty-four studies in the U.S.A, Europe, Israel and Asia were included. In our main analysis, we excluded patients who were known or who were likely to have not received platinum-based therapy. Compared to non-carriers, BRCA1 and BRCA2 carriers were more likely to present with advanced stage (BRCA1; p=2×10−4, BRCA2; p=4×10−6), high grade (BRCA1; p=4×10−9, BRCA2; p=1×10−4), serous disease (BRCA1; p=3×10−6, BRCA2; p=0.003). BRCA1 carriers were younger at diagnosis (p=2×10−9) and BRCA2 slightly older at diagnosis (p=0.002) than non-carriers. BRCA1 and BRCA2 carriers did not differ in terms of tumor stage, grade or histology but did show differences in the age at diagnosis (p=2×10−14). In an unadjusted analysis, neither BRCA1 nor BRCA2 carriers differed from non-carriers in overall survival (BRCA1: HR=1.02, p=0.77, BRCA2: HR=0.89, p=0.36). After adjusting for stage, grade, histology and age at diagnosis, BRCA1 carriers showed a modest, non-significant, survival advantage (HR=0.84, p=0.12) while BRCA2 carriers showed a marked improvement (HR=0.57, p=6×10−4) compared to non-carriers. This survival advantage was also seen when comparing BRCA2 carriers to BRCA1 carriers after adjustment for age at diagnosis (HR=0.69, p=5×10−4). Similar results were obtained when we restricted the analysis to high grade, advanced stage, serous cases. The survival differences we observed between BRCA1 and BRCA2 carriers could be related to differences in tumor biology or chemo-sensitivity. Additional analyses investigating the later will be presented in addition to mutation-specific effects within BRCA1 and BRCA2. In this study, we provide strong evidence for genetic heterogeneity of survival effects in ovarian cancer patients with BRCA mutations. This could be applied to both clinical prediction and to aid our understanding of the complex biology of BRCA mutations. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 2752. doi:10.1158/1538-7445.AM2011-2752

The paper develops a genetic algorithm (GA) based "growing" technique to design and synthesise analogue circuits with practical constraints, such as the manufacturer's preferred component values. Most existing problems when evolutionary search techniques are applied to circuit design are addressed. The developed GA technique is then applied both to synthesise the topology of a network and perform value optimisation on the components based on a set of commonly used component values (E-12 series). Passive filter networks synthesised this way are realisable, effective and of novel topology. It is anticipated that this technique can be extended to active networks.