Transparent and autonomic rollback-recovery in cluster systems

Cluster systems provide an excellent environment to run computation hungry applications. However, due to being created using commodity components they are prone to failures. To overcome these failures we propose to use rollback-recovery, which consists of the checkpointing and recovery facilities. Checkpointing facilities have been the focus of many previous studies; however, the recovery facilities have been overlooked. This paper focuses on the requirements, concept and architecture of recovery facilities. The synthesized fault tolerant system was implemented in the GENESIS system and evaluated. The results show that the synthesized system is efficient and scalable.

Texture recognition by using GLCM and various aggregation functions

We discuss the problem of texture recognition based on the grey level co-occurrence matrix (GLCM). We performed a number of numerical experiments to establish whether the accuracy of classification is optimal when GLCM entries are aggregated into standard metrics like contrast, dissimilarity, homogeneity, entropy, etc., and compared these metrics to several alternative aggregation methods.We conclude that k nearest neighbors classification based on raw GLCM entries typically works better than classification based on the standard metrics for noiseless data, that metrics based on principal component analysis inprove classification, and that a simple change from the arithmetic to quadratic mean in calculating the standard metrics also improves classification.

Inhibition of platelet aggregation by omega-3 polyunsaturated fatty acids is gender specific — redefining platelet response to fish oils

Existence of gender differences in cardiovascular disease (CVD) following long-chain omega-3 polyunsaturated fatty acid (LCn-3 PUFA) supplementation have suggested that sex hormones play a role in cardio-protection. The objective of this study was to determine gender specific responses in the efficacy of LCn-3 PUFA to inhibit platelet aggregation in vitro. Blood was analyzed for collagen-induced platelet aggregation following pre-incubation with LCn-3 PUFA in healthy adults (n=42). Eicosapentaenoic acid (EPA) was significantly more effective in reducing platelet aggregation compared with docosapentaenoic acid (DPA) and docosahexaenoic acid (DHA). When grouped by gender, this differential pattern was followed in males only. In females, DHA, DPA and EPA were all equally effective. Between group analyses (LCn-3 PUFA vs. gender) showed that both DHA and DPA were significantly less effective in males compared with females. EPA was equally effective in reducing platelet aggregation in both groups. These findings show that significant gender differences exist in platelet aggregation in response to various LCn-3 PUFA treatments.

Adolescents' use of purpose built shade in secondary schools : cluster randomised controlled trial

Objective To examine whether students use or avoid newly shaded areas created by shade sails installed at schools.

Design Cluster randomised controlled trial with secondary schools as the unit of randomisation.

Setting 51 secondary schools with limited available shade, in Australia, assessed over two spring and summer terms.

Participants Students outside at lunch times.

Intervention Purpose built shade sails were installed in winter 2005 at full sun study sites to increase available shade for students in the school grounds.

Main outcome measure Mean number of students using the primary study sites during weekly observations at lunch time.

Results Over the study period the mean change in students using the primary study site from pre-test to post-test was 2.63 (95% confidence interval 0.87 to 4.39) students in intervention schools and –0.03 (–1.16 to 1.09) students in control schools. The difference in mean change between groups was 2.67 (0.65 to 4.68) students (P=0.011).

Conclusions Students used rather than avoided newly shaded areas provided by purpose built shade sails at secondary schools in this trial, suggesting a practical means of reducing adolescents’ exposure to ultraviolet radiation.

Trial registration Exempt.

MOSAIC (MOthers' Advocates In the Community): protocol and sample description of a cluster randomised trial of mentor mother support to reduce intimate partner violence among pregnant or recent mothers

Background : Intimate partner violence (IPV) is prevalent globally, experienced by a significant minority of women in the early childbearing years and is harmful to the mental and physical health of women and children. There are very few studies with rigorous designs which have tested the effectiveness of IPV interventions to improve the health and wellbeing of abused women. Evidence for the separate benefit to victims of social support, advocacy and non-professional mentoring suggested that a combined model may reduce the levels of violence, the associated mental health damage and may increase a woman's health, safety and connection with her children. This paper describes the development, design and implementation of a trial of mentor mother support set in primary care, including baseline characteristics of participating women.

Methods/Design : MOSAIC (MOtherS' Advocates In the Community) was a cluster randomised trial embedded in general practice and maternal and child health (MCH) nursing services in disadvantaged suburbs of Melbourne, Australia. Women who were pregnant or with infants, identified as abused or symptomatic of abuse, were referred by IPV-trained GPs and MCH nurses from 24 general practices and eight nurse teams from January 2006 to December 2007. Women in the intervention arm received up to 12 months support from trained and supported non-professional mentor mothers. Vietnamese health professionals also referred Vietnamese women to bilingual mentors in a sub-study. Baseline and follow-up surveys at 12 months measured IPV (CAS), depression (EPDS), general health (SF-36), social support (MOS-SF) and attachment to children (PSI-SF). Significant development and piloting occurred prior to trial commencement. Implementation interviews with MCH nurses, GPs and mentors assisted further refinement of the intervention. In-depth interviews with participants and mentors, and follow-up surveys of MCH nurses and GPs at trial conclusion will shed further light on MOSAIC's impact.

Discussion : Despite significant challenges, MOSAIC will make an important contribution to the need for evidence of effective partner violence interventions, the role of non-professional mentors in partner violence support services and the need for more evaluation of effective health professional training and support in caring for abused women and children among their populations.

Weight, physical activity and dietary behavior change in young mothers : short term results of the HeLP-her cluster randomized controlled trial

Background
Preventing weight gain rather than treating established obesity is an important economic and public health response to the rapidly increasing rates of obesity worldwide. Treatment of established obesity is complex and costly requiring multiple resources. Preventing weight gain potentially requires fewer resources to reach broad population groups, yet there is little evidence for successful interventions to prevent weight gain in the community. Women with children are an important target group because of high rates of weight gain and the potential to influence the health behaviors in family members.

Methods
The aim of this cluster randomized controlled trial was to evaluate the short term effect of a community-based self-management intervention to prevent weight gain. Two hundred and fifty mothers of young children (mean age 40 years ± 4.5, BMI 27.9 kg/m2 ± 5.6) were recruited from the community in Melbourne, Australia. The intervention group (n = 127) attended four interactive group sessions over 4 months, held in 12 local primary schools in 2006, and was compared to a group (n = 123) receiving a single, non-interactive, health education session. Data collection included self-reported weight (both groups), measured weight (intervention only), self-efficacy, dietary intake and physical activity.

Results
Mean measured weight decreased significantly in the intervention group (-0.78 kg 95% CI; -1.22 to -0.34, p < 0.001). Comparing groups using self-reported weight, both the intervention and comparison groups decreased weight, -0.75 kg (95% CI; -1.57 to 0.07, p = 0.07) and -0.72 kg (95% CI; -1.59 to 0.14 p = 0.10) respectively with no significant difference between groups (-0.03 kg, 95% CI; -1.32 to 1.26, p = 0.95). More women lost or maintained weight in the intervention group. The intervention group tended to have the greatest effect in those who were overweight at baseline and in those who weighed themselves regularly. Intervention women who rarely self-weighed gained weight (+0.07 kg) and regular self-weighers lost weight (-1.66 kg) a difference of -1.73 kg (95% CI; -3.35 to -0.11 p = 0.04). The intervention reported increased physical activity although the difference between groups did not reach significance. Both groups reported replacing high fat foods with low fat alternatives and self-efficacy deteriorated in the comparison group only.

Conclusion
Both a single health education session and interactive behavioral intervention will result in a similar weight loss in the short term, although more participants in the interactive intervention lost or maintained weight. There were small non-significant changes to physical activity and changes to fat intake specifically replacing high fat foods with low fat alternatives such as fruit and vegetables. Self-monitoring appears to enhance weight loss when part of an intervention.

Evaluation of a decision aid for prenatal testing of fetal abnormalities : a cluster randomised trial [ISRCTN22532458]

Background: By providing information on the relative merits and potential harms of the options available and a framework to clarify preferences, decision aids can improve knowledge and realistic expectations and decrease decisional conflict in individuals facing decisions between alternative forms of action. Decision-making about prenatal testing for fetal abnormalities is often confusing and difficult for women and the effectiveness of decision aids in this field has not been established. This study aims to test whether a decision aid for prenatal testing of fetal abnormalities, when compared to a pamphlet, improves women's informed decision-making and decreases decisional conflict.

Methods/design: A cluster designed randomised controlled trial is being conducted in Victoria, Australia. Fifty General Practitioners (GPs) have been randomised to one of two arms: providing women with either a decision aid or a pamphlet. The two primary outcomes will be measured by comparing the difference in percentages of women identified as making an informed choice and the difference in mean decisional conflict scores between the two groups. Data will be collected from women using questionnaires at 14 weeks and 24 weeks gestation.

The sample size of 159 women in both arms of the trial has been calculated to detect a difference of 18% (50 to 68%) in informed choice between the two groups. The required numbers have been adjusted to accommodate the cluster design, miscarriage and participant lost – to – follow up.

Baseline characteristics of women will be summarised for both arms of the trial. Similarly, characteristics of GPs will be compared between arms.

Differences in the primary outcomes will be analysed using 'intention-to-treat' principles. Appropriate regression techniques will adjust for the effects of clustering and include covariates to adjust for the stratifying variable and major potential confounding factors.

Discussion: The findings from this trial will make a significant contribution to improving women's experience of prenatal testing and will have application to a variety of maternity care settings. The evaluation of a tailored decision aid will also have implications for pregnancy care providers by identifying whether or not such a resource will support their role in providing prenatal testing information.

Performance execution of multiple NAS parallel programs on a cluster

Currently, coordinated scheduling of multiple parallel applications across computers has been considered as the critical factor to achieve high execution performance. We claim in this report that the performance and costs of the execution of parallel applications could be improved if not only dedicated clusters but also non-dedicated clusters were used and several parallel applications were executed concurreontly. To support this claim we carried out experimental study into the performance of multiple NAS parallel programs executing concurrently on a non-dedicated cluster.

Scheduling of a parallel computation-bound application and sequential applications executing concurrently on a cluster - a case study.

Studies have shown that most of the computers in a non-dedicated cluster are often idle or lightly loaded. The underutilized computers in a non-dedicated cluster can be employed to execute parallel applications. The aim of this study is to learn how concurrent execution of a computation-bound and sequential applications influence their execution performance and cluster utilization. The result of the study has demonstrated that a computation-bound parallel application benefits from load balancing, and at the same time sequential applications suffer only an insignificant slowdown of execution. Overall, the utilization of a non-dedicated cluster is improved.

Some aspects of adaptive cluster sampling

The adaptive cluster sampling (ACS) procedure is difficult to apply if some of the networks appearing in the sample are large. To deal with such large networks, a two-stage adaptive cluster sampling (TACS) procedure and an adjusted two-stage adaptive cluster sampling (ATACS) procedure are discussed.