Dynamic load balancing in parallel KD-tree k-means

One among the most influential and popular data mining methods is the k-Means algorithm for cluster analysis. Techniques for improving the efficiency of k-Means have been largely explored in two main directions. The amount of computation can be significantly reduced by adopting geometrical constraints and an efficient data structure, notably a multidimensional binary search tree (KD-Tree). These techniques allow to reduce the number of distance computations the algorithm performs at each iteration. A second direction is parallel processing, where data and computation loads are distributed over many processing nodes. However, little work has been done to provide a parallel formulation of the efficient sequential techniques based on KD-Trees. Such approaches are expected to have an irregular distribution of computation load and can suffer from load imbalance. This issue has so far limited the adoption of these efficient k-Means variants in parallel computing environments. In this work, we provide a parallel formulation of the KD-Tree based k-Means algorithm for distributed memory systems and address its load balancing issue. Three solutions have been developed and tested. Two approaches are based on a static partitioning of the data set and a third solution incorporates a dynamic load balancing policy.

Reductions in insecticide use from adoption of Bt cotton in South Africa: impacts on economic performance and toxic load to the environment

The study reported presents the findings relating to commercial growing of genetically-modified Bt cotton in South Africa by a large sample of smallholder farmers over three seasons (1998/99, 1999/2000, 2000/01) following adoption. The analysis presents constructs and compares groupwise differences for key variables in Bt v. non-Bt technology and uses regressions to further analyse the production and profit impacts of Bt adoption. Analysis of the distribution of benefits between farmers due to the technology is also presented. In parallel with these socio-economic measures, the toxic loads being presented to the environment following the introduction of Bt cotton are monitored in terms of insecticide active ingredient (ai) and the Biocide Index. The latter adjusts ai to allow for differing persistence and toxicity of insecticides. Results show substantial and significant financial benefits to smallholder cotton growers of adopting Bt cotton over three seasons in terms of increased yields, lower insecticide spray costs and higher gross margins. This includes one particularly wet, poor growing season. In addition, those with the smaller holdings appeared to benefit proportionately more from the technology (in terms of higher gross margins) than those with larger holdings. Analysis using the Gini-coefficient suggests that the Bt technology has helped to reduce inequality amongst smallholder cotton growers in Makhathini compared to what may have been the position if they had grown conventional cotton. However, while Bt growers applied lower amounts of insecticide and had lower Biocide Indices (per ha) than growers of non-Bt cotton, some of this advantage was due to a reduction in non-bollworm insecticide. Indeed, the Biocide Index for all farmers in the population actually increased with the introduction of Bt cotton. The results indicate the complexity of such studies on the socio-economic and environmental impacts of GM varieties in the developing world.

'A load too heavy': Critical reflections on the child labor problem in Africa's small-scale mining sector

This paper provides an extended analysis of the child labor problem in the artisanal and small-scale mining (ASM) sector, focusing specifically on the situation in sub-Saharan Africa. In recent years, the issue of child labor in ASM has garnered significant attention from the International Labor Organization (ILO), which has been particularly active in raising public awareness of the problem; and, has proceeded to implement policies and collaborative project work aimed at Curtailing children's participation in ASM activities in a number of African countries. The analysis concludes with a critical appraisal of an ILO project recently launched in the Talensi-Nabdam District in the Upper East Region of Ghana, which sheds light on how the child labor problem is being tackled in practice in ASM communities in sub-Saharan Africa. (c) 2008 Elsevier Ltd. All rights reserved.

The failure of meta-analytic asymptotics for the seemingly efficient estimator of the common risk difference

We consider the case of a multicenter trial in which the center specific sample sizes are potentially small. Under homogeneity, the conventional procedure is to pool information using a weighted estimator where the weights used are inverse estimated center-specific variances. Whereas this procedure is efficient for conventional asymptotics (e. g. center-specific sample sizes become large, number of center fixed), it is commonly believed that the efficiency of this estimator holds true also for meta-analytic asymptotics (e.g. center-specific sample size bounded, potentially small, and number of centers large). In this contribution we demonstrate that this estimator fails to be efficient. In fact, it shows a persistent bias with increasing number of centers showing that it isnot meta-consistent. In addition, we show that the Cochran and Mantel-Haenszel weighted estimators are meta-consistent and, in more generality, provide conditions on the weights such that the associated weighted estimator is meta-consistent.

A jackknife variance estimator for unequal probability sampling

The jackknife method is often used for variance estimation in sample surveys but has only been developed for a limited class of sampling designs.We propose a jackknife variance estimator which is defined for any without-replacement unequal probability sampling design. We demonstrate design consistency of this estimator for a broad class of point estimators. A Monte Carlo study shows how the proposed estimator may improve on existing estimators.

Effect of viral load on the outcome of herpes zoster

Varicella-zoster virus (VZV) is a member of the Herpesviridae family, primary infection with which causes varicella, more commonly known as chicken pox. Characteristic of members of the alphaherpesvirus subfamily, VZV is neurotropic and establishes latency in sensory neurons. Reactivation of VZV causes herpes zoster, also known as shingles. The most frequent complication following zoster is chronic and often debilitating pain called postherpetic neuralgia (PHN), which can last for months after the disappearance of a rash. During episodes of acute zoster, VZV viremia occurs in some, but not all, patients; however, the effect of the viral load on the disease outcome is not known. Here we describe the development of a highly specific, sensitive, and reproducible real-time PCR assay to investigate the factors that may contribute to the presence and levels of baseline viremia in patients with zoster and to determine the relationship between viremia and the development and persistence of PHN. VZV DNA was detected in the peripheral blood mononuclear cells (PBMCs) of 78% of patients with acute zoster and in 9% of healthy asymptomatic blood donors. The presence of VZV in the PBMCs of patients with acute zoster was independently associated with age and being on antivirals but not with gender, immune status, extent of rash, the age of the rash at the time of blood sampling, having a history of prodromal pain, or the extent of acute pain. Prodromal pain was significantly associated with higher baseline viral loads. Viral load levels were not associated with the development or persistence of PHN at 6, 12, or 26 weeks.

Comparative evaluation of a new effective population size estimator based on approximate Bayesian computation

We describe and evaluate a new estimator of the effective population size (N-e), a critical parameter in evolutionary and conservation biology. This new "SummStat" N-e. estimator is based upon the use of summary statistics in an approximate Bayesian computation framework to infer N-e. Simulations of a Wright-Fisher population with known N-e show that the SummStat estimator is useful across a realistic range of individuals and loci sampled, generations between samples, and N-e values. We also address the paucity of information about the relative performance of N-e estimators by comparing the SUMMStat estimator to two recently developed likelihood-based estimators and a traditional moment-based estimator. The SummStat estimator is the least biased of the four estimators compared. In 32 of 36 parameter combinations investigated rising initial allele frequencies drawn from a Dirichlet distribution, it has the lowest bias. The relative mean square error (RMSE) of the SummStat estimator was generally intermediate to the others. All of the estimators had RMSE > 1 when small samples (n = 20, five loci) were collected a generation apart. In contrast, when samples were separated by three or more generations and Ne less than or equal to 50, the SummStat and likelihood-based estimators all had greatly reduced RMSE. Under the conditions simulated, SummStat confidence intervals were more conservative than the likelihood-based estimators and more likely to include true N-e. The greatest strength of the SummStat estimator is its flexible structure. This flexibility allows it to incorporate any, potentially informative summary statistic from Population genetic data.

Coxsackievirus B3-associated myocardial pathology and viral load reduced by recombinant soluble human decay-accelerating factor in mice

Coxsackievirus B3 (CVB3) infection can result in myocarditis, which in turn may lead to a protracted immune response and subsequent dilated cardiomyopathy. Human decay-accelerating factor (DAF), a binding receptor for CVB3, was synthesized as a soluble IgG1-Fc fusion protein (DAF-Fc). In vitro, DAF-Fc was able to inhibit complement activity and block infection by CVB3, although blockade of infection varied widely among strains of CVB3. To determine the effects of DAF-Fc in vivo, 40 adolescent A/J mice were infected with a myopathic strain of CVB3 and given DAF-Fc treatment 3 days before infection, during infection, or 3 days after infection; the mice were compared with virus alone and sham-infected animals. Sections of heart, spleen, kidney, pancreas, and liver were stained with hematoxylin and eosin and submitted to in situ hybridization for both positive-strand and negative-strand viral RNA to determine the extent of myocarditis and viral infection, respectively. Salient histopathologic features, including myocardial lesion area, cell death, calcification and inflammatory cell infiltration, pancreatitis, and hepatitis were scored without knowledge of the experimental groups. DAF-Fc treatment of mice either preceding or concurrent with CVB3 infection resulted in a significant decrease in myocardial lesion area and cell death and a reduction in the presence of viral RNA. All DAF-Fc treatment groups had reduced infectious CVB3 recoverable from the heart after infection. DAF-Fc may be a novel therapeutic agent for active myocarditis and acute dilated cardiomyopathy if given early in the infectious period, although more studies are needed to determine its mechanism and efficacy.

Sparse kernel density estimator using orthogonal regression based on D-optimality experimental design

A novel sparse kernel density estimator is derived based on a regression approach, which selects a very small subset of significant kernels by means of the D-optimality experimental design criterion using an orthogonal forward selection procedure. The weights of the resulting sparse kernel model are calculated using the multiplicative nonnegative quadratic programming algorithm. The proposed method is computationally attractive, in comparison with many existing kernel density estimation algorithms. Our numerical results also show that the proposed method compares favourably with other existing methods, in terms of both test accuracy and model sparsity, for constructing kernel density estimates.

M-estimator, and D-optimality model construction using orthogonal forward regression

This correspondence introduces a new orthogonal forward regression (OFR) model identification algorithm using D-optimality for model structure selection and is based on an M-estimators of parameter estimates. M-estimator is a classical robust parameter estimation technique to tackle bad data conditions such as outliers. Computationally, The M-estimator can be derived using an iterative reweighted least squares (IRLS) algorithm. D-optimality is a model structure robustness criterion in experimental design to tackle ill-conditioning in model Structure. The orthogonal forward regression (OFR), often based on the modified Gram-Schmidt procedure, is an efficient method incorporating structure selection and parameter estimation simultaneously. The basic idea of the proposed approach is to incorporate an IRLS inner loop into the modified Gram-Schmidt procedure. In this manner, the OFR algorithm for parsimonious model structure determination is extended to bad data conditions with improved performance via the derivation of parameter M-estimators with inherent robustness to outliers. Numerical examples are included to demonstrate the effectiveness of the proposed algorithm.

A bagging-based correction for the mixture model estimator of population size

Estimation of a population size by means of capture-recapture techniques is an important problem occurring in many areas of life and social sciences. We consider the frequencies of frequencies situation, where a count variable is used to summarize how often a unit has been identified in the target population of interest. The distribution of this count variable is zero-truncated since zero identifications do not occur in the sample. As an application we consider the surveillance of scrapie in Great Britain. In this case study holdings with scrapie that are not identified (zero counts) do not enter the surveillance database. The count variable of interest is the number of scrapie cases per holding. For count distributions a common model is the Poisson distribution and, to adjust for potential heterogeneity, a discrete mixture of Poisson distributions is used. Mixtures of Poissons usually provide an excellent fit as will be demonstrated in the application of interest. However, as it has been recently demonstrated, mixtures also suffer under the so-called boundary problem, resulting in overestimation of population size. It is suggested here to select the mixture model on the basis of the Bayesian Information Criterion. This strategy is further refined by employing a bagging procedure leading to a series of estimates of population size. Using the median of this series, highly influential size estimates are avoided. In limited simulation studies it is shown that the procedure leads to estimates with remarkable small bias.

Quantification of mitochondrial DNA mutation load

Mitochondrial DNA (mtDNA) mutations are an important cause of genetic disease and have been proposed to play a role in the ageing process. Quantification of total mtDNA mutation load in ageing tissues is difficult as mutational events are rare in a background of wild-type molecules, and detection of individual mutated molecules is beyond the sensitivity of most sequencing based techniques. The methods currently most commonly used to document the incidence of mtDNA point mutations in ageing include post-PCR cloning, single-molecule PCR and the random mutation capture assay. The mtDNA mutation load obtained by these different techniques varies by orders of magnitude, but direct comparison of the three techniques on the same ageing human tissue has not been performed. We assess the procedures and practicalities involved in each of these three assays and discuss the results obtained by investigation of mutation loads in colonic mucosal biopsies from ten human subjects.

Part load operation coefficient of air-conditioning system of public building

IPLV overall coefficient, presented by Air-Conditioning and Refrigeration Institute (ARI) of America, shows running/operation status of air-conditioning system host only. For overall operation coefficient, logical solution has not been developed, to reflect the whole air-conditioning system under part load. In this research undertaking, the running time proportions of air-conditioning systems under part load have been obtained through analysis on energy consumption data during practical operation in all public buildings in Chongqing. This was achieved by using analysis methods, based on the statistical energy consumption data distribution of public buildings month-by-month. Comparing with the weight number of IPLV, part load operation coefficient of air-conditioning system, based on this research, does not only show the status of system refrigerating host, but also reflects and calculate energy efficiency of the whole air-conditioning system. The coefficient results from the processing and analyzing of practical running data, shows the practical running status of area and building type (actual and objective) – not clear. The method is different from model analysis which gets IPLV weight number, in the sense that this method of coefficient results in both four equal proportions and also part load operation coefficient of air-conditioning system under any load rate as necessary.

Minimum neighbour and extended Kalman filter estimator: a practical distributed channel assignment scheme for dense wireless local area networks

Dense deployments of wireless local area networks (WLANs) are becoming a norm in many cities around the world. However, increased interference and traffic demands can severely limit the aggregate throughput achievable unless an effective channel assignment scheme is used. In this work, a simple and effective distributed channel assignment (DCA) scheme is proposed. It is shown that in order to maximise throughput, each access point (AP) simply chooses the channel with the minimum number of active neighbour nodes (i.e. nodes associated with neighbouring APs that have packets to send). However, application of such a scheme to practice depends critically on its ability to estimate the number of neighbour nodes in each channel, for which no practical estimator has been proposed before. In view of this, an extended Kalman filter (EKF) estimator and an estimate of the number of nodes by AP are proposed. These not only provide fast and accurate estimates but can also exploit channel switching information of neighbouring APs. Extensive packet level simulation results show that the proposed minimum neighbour and EKF estimator (MINEK) scheme is highly scalable and can provide significant throughput improvement over other channel assignment schemes.