A Unified Scheduler for Recursive and Task Dataflow Parallelism

Task dataflow languages simplify the specification of parallel programs by dynamically detecting and enforcing dependencies between tasks. These languages are, however, often restricted to a single level of parallelism. This language design is reflected in the runtime system, where a master thread explicitly generates a task graph and worker threads execute ready tasks and wake-up their dependents. Such an approach is incompatible with state-of-the-art schedulers such as the Cilk scheduler, that minimize the creation of idle tasks (work-first principle) and place all task creation and scheduling off the critical path. This paper proposes an extension to the Cilk scheduler in order to reconcile task dependencies with the work-first principle. We discuss the impact of task dependencies on the properties of the Cilk scheduler. Furthermore, we propose a low-overhead ticket-based technique for dependency tracking and enforcement at the object level. Our scheduler also supports renaming of objects in order to increase task-level parallelism. Renaming is implemented using versioned objects, a new type of hyper object. Experimental evaluation shows that the unified scheduler is as efficient as the Cilk scheduler when tasks have no dependencies. Moreover, the unified scheduler is more efficient than SMPSS, a particular implementation of a task dataflow language.

Behavioural and cognitive-behavioural group-based parenting programmes for early-onset conduct problems in children aged 3 to 12 years.

Early-onset child conduct problems are common and costly. A large number of studies and some previous reviews have focused on behavioural and cognitive-behavioural group-based parenting interventions, but methodological limitations are commonplace and evidence for the effectiveness and cost-effectiveness of these programmes has been unclear. To assess the effectiveness and cost-effectiveness of behavioural and cognitive-behavioural group-based parenting programmes for improving child conduct problems, parental mental health and parenting skills. We searched the following databases between 23 and 31 January 2011: CENTRAL (2011, Issue 1), MEDLINE (1950 to current), EMBASE (1980 to current), CINAHL (1982 to current), PsycINFO (1872 to current), Social Science Citation Index (1956 to current), ASSIA (1987 to current), ERIC (1966 to current), Sociological Abstracts (1963 to current), Academic Search Premier (1970 to current), Econlit (1969 to current), PEDE (1980 to current), Dissertations and Theses Abstracts (1980 to present), NHS EED (searched 31 January 2011), HEED (searched 31 January 2011), DARE (searched 31 January 2011), HTA (searched 31 January 2011), mRCT (searched 29 January 2011). We searched the following parent training websites on 31 January 2011: Triple P Library, Incredible Years Library and Parent Management Training. We also searched the reference lists of studies and reviews. We included studies if: (1) they involved randomised controlled trials (RCTs) or quasi-randomised controlled trials of behavioural and cognitive-behavioural group-based parenting interventions for parents of children aged 3 to 12 years with conduct problems, and (2) incorporated an intervention group versus a waiting list, no treatment or standard treatment control group. We only included studies that used at least one standardised instrument to measure child conduct problems. Two authors independently assessed the risk of bias in the trials and the methodological quality of health economic studies. Two authors also independently extracted data. We contacted study authors for additional information. This review includes 13 trials (10 RCTs and three quasi-randomised trials), as well as two economic evaluations based on two of the trials. Overall, there were 1078 participants (646 in the intervention group; 432 in the control group). The results indicate that parent training produced a statistically significant reduction in child conduct problems, whether assessed by parents (standardised mean difference (SMD) -0.53; 95% confidence interval (CI) -0.72 to -0.34) or independently assessed (SMD -0.44; 95% CI -0.77 to -0.11). The intervention led to statistically significant improvements in parental mental health (SMD -0.36; 95% CI -0.52 to -0.20) and positive parenting skills, based on both parent reports (SMD -0.53; 95% CI -0.90 to -0.16) and independent reports (SMD -0.47; 95% CI -0.65 to -0.29). Parent training also produced a statistically significant reduction in negative or harsh parenting practices according to both parent reports (SMD -0.77; 95% CI -0.96 to -0.59) and independent assessments (SMD -0.42; 95% CI -0.67 to -0.16). Moreover, the intervention demonstrated evidence of cost-effectiveness. When compared to a waiting list control group, there was a cost of approximately $2500 (GBP 1712; EUR 2217) per family to bring the average child with clinical levels of conduct problems into the non-clinical range. These costs of programme delivery are modest when compared with the long-term health, social, educational and legal costs associated with childhood conduct problems. Behavioural and cognitive-behavioural group-based parenting interventions are effective and cost-effective for improving child conduct problems, parental mental health and parenting skills in the short term. The cost of programme delivery was modest when compared with the long-term health, social, educational and legal costs associated with childhood conduct problems. Further research is needed on the long-term assessment of outcomes.

An Investigation Into Features For Multi-View Lipreading

For the first time in this paper we present results showing the effect of speaker head pose angle on automatic lip-reading performance over a wide range of closely spaced angles. We analyse the effect head pose has upon the features themselves and show that by selecting coefficients with minimum variance w.r.t. pose angle, recognition performance can be improved when train-test pose angles differ. Experiments are conducted using the initial phase of a unique multi view Audio-Visual database designed specifically for research and development of pose-invariant lip-reading systems. We firstly show that it is the higher order horizontal spatial frequency components that become most detrimental as the pose deviates. Secondly we assess the performance of different feature selection masks across a range of pose angles including a new mask based on Minimum Cross-Pose Variance coefficients. We report a relative improvement of 50% in Word Error Rate when using our selection mask over a common energy based selection during profile view lip-reading.

Indigenous Arbuscular Mycorrhizal Fungal Assemblages Protect Grassland Host Plants from Pathogens

Plant roots can establish associations with neutral, beneficial and pathogenic groups of soil organisms. Although it has been recognized from the study of individual isolates that these associations are individually important for plant growth, little is known about interactions of whole assemblages of beneficial and pathogenic microorganisms associating with plants. We investigated the influence of an interaction between local arbuscular mycorrhizal (AM) fungal and pathogenic/saprobic microbial assemblages on the growth of two different plant species from semi-arid grasslands in NE Germany (Mallnow near Berlin). In a greenhouse experiment each plant species was grown for six months in either sterile soil or in sterile soil with one of three different treatments: 1) an AM fungal spore fraction isolated from field soil from Mallnow; 2) a soil pathogen/saprobe fraction consisting of a microbial community prepared with field soil from Mallnow and; 3) the combined AM fungal and pathogen/saprobe fractions. While both plant species grew significantly larger in the presence of AM fungi, they responded negatively to the pathogen/saprobe treatment. For both plant species, we found evidence of pathogen protection effects provided by the AM fungal assemblages. These results indicate that interactions between assemblages of beneficial and pathogenic microorganisms can influence the growth of host plants, but that the magnitude of these effects is plant species-specific.