Standard Bag Research Articles

BAGGED ONE-CLASS CLASSIFIERS IN THE PRESENCE OF OUTLIERS

The problem of training classifiers only with target data arises in many applications where nontarget data are too costly, difficult to obtain, or not available at all. Several one-class classification methods have been presented to solve this problem, but most of the methods are highly sensitive to the presence of outliers in the target class. Ensemble methods have therefore been proposed as a powerful way to improve the classification performance of binary/multi-class learning algorithms by introducing diversity into classifiers. However, their application to one-class classification has been rather limited. In this paper, we present a new ensemble method based on a nonparametric weighted bagging strategy for one-class classification, to improve accuracy in the presence of outliers. While the standard bagging strategy assumes a uniform data distribution, the method we propose here estimates a probability density based on a forest structure of the data. This assumption allows the estimation of data distribution from the computation of simple univariate and bivariate kernel densities. Experiments using original and noisy versions of 20 different datasets show that bagging ensemble methods applied to different one-class classifiers outperform base one-class classification methods. Moreover, we show that, in noisy versions of the datasets, the nonparametric weighted bagging strategy we propose outperforms the classical bagging strategy in a statistically significant way.

Comparison of standard versus tailored parenteral nutrition in paediatric intensive care

ObjectivePatients in paediatric intensive care (PICU) are often fluid restricted and require a high volume of intravenous infusions, allowing little volume for nutritional intake. Bespoke parenteral nutrition (TPN) is associated...

Hybrid Ensemble Construction with Selected Neural Networks

A Neural Network Ensemble (NNE) is convenient for improving classification task performance. Among the remarkable number of methods based on different techniques for constructing NNEs, Negative Correlation Learning (NCL), bagging, and boosting are the most popular. None of them, however, could show better performance for all problems. To improve performance combining the complementary strengths of the individual methods, we propose two different ways to construct hybrid ensembles combining NCL with bagging and boosting. One produces a pool of predefined numbers of networks using standard NCL and bagging (or boosting) and then uses a genetic algorithm to select an optimal network subset for an NNE from the pool. Results of experiments confirmed that our proposals show consistently better performance with concise ensembles than conventional methods when tested using a suite of 25 benchmark problems.

Review of standard & individualised parenteral nutrition (PN) prescribing in neonates

ObjectivesPrimarily, the review was carried out to look at the appropriateness of parenteral nutrition (PN) prescribing in neonates against ESPGHAN guidelines using both standard and individually manufactured bags.1 Secondly, to...

Excreta disposal in emergencies: Bag and Peepoo trials with internally displaced people in Port-au-Prince

After a series of earthquakes devastated Port-au-Prince, Haiti, on 12 January 2010, safe excreta disposal became an urgent priority. To mainstream innovative approaches to sanitation within the realities of urban humanitarian response, Oxfam GB undertook a trial from April to May 2010, of standard bag and Peepoo excreta disposal systems in two IDP settlements. Trial results demonstrate that with proper collection and removal, both bags and Peepoos are viable excreta disposal options in emergencies. A reduction was seen in ‘flying toilets’, open defecation and user reports on diarrhoea in both settlements. For in-home use, the Peepoo was preferred over bags, 84 per cent to 55 per cent (χ2 = 6.15, p=0.013) based on its ability to contain odour. Key programmatic recommendations include contingency planning for disaster-prone areas and further analysis of cost effectiveness and phase-out points for bag excreta disposal systems. Additionally, the Peepoo size should be adjusted to be more widely applicable in diverse emergency settings.

Effective Spore Density of <i>Glomus mosseae</i>, Arbuscular Mycorrhiza (AM), for Inoculation of Rooted Cuttings of Black Pepper (<i>Piper nigrum</i> Linn.)

Introduction of mycorrhizal associations is a complementary alternative to improve fertilizer absorption efficiency of black pepper ( Piper nigrum Linn.). Therefore, as an initial step, this experiment was carried out to select a suitable spore density of the Arbuscular Mycorrhize (AM) fungus Glomus mosseae for inoculation of rooted cuttings of black pepper in the nursery. Four mycorrhizal inoculum levels of G. mosseae , namely 25 g (T1), 75 g (T2), 150 g (T3) and 300 g (T4) were tested with a control (T5) after quantification of the initial spore density values. The respective quantities of inoculum were incorporated separately to standard size polythene bags (21 cm diameter x 13 cm height; gauge 150) filled with a sterilized standard potting mixture. A two nodal cutting of black pepper (local selection GK 49) was planted in each polythene bag and weeding and watering were done as required. Observation of darkly stained AM hyphae at the 2nd month after inoculation in all the inoculated treatments confirmed the success of inoculation. The heaviest infection of mycorrhizal fungi was observed in roots of T3 treatment showing root colonization, along with mycelium (100 %), vesicles (94 %), and spores (769/50 g) at the 6th month after inoculation. Significantly higher (p DOI: 10.4038/tar.v21i2.2600 Tropical Agricultural Research Vol. 21(2): 189-197 (2009)

A novel ensemble construction method for multi-view data using random cross-view correlation between within-class examples

Correlated information between multiple views can provide useful information for building robust classifiers. One way to extract correlated features from different views is using canonical correlation analysis (CCA). However, CCA is an unsupervised method and can not preserve discriminant information in feature extraction. In this paper, we first incorporate discriminant information into CCA by using random cross-view correlations between within-class examples. Because of the random property, we can construct a lot of feature extractors based on CCA and random correlation. So furthermore, we fuse those feature extractors and propose a novel method called random correlation ensemble (RCE) for multi-view ensemble learning. We compare RCE with existing multi-view feature extraction methods including CCA and discriminant CCA (DCCA) which use all cross-view correlations between within-class examples, as well as the trivial ensembles of CCA and DCCA which adopt standard bagging and boosting strategies for ensemble learning. Experimental results on several multi-view data sets validate the effectiveness of the proposed method.

Evaluation of Two Different Hand Hygiene Procedures during Routine Patient Care

In this study, the antimicrobial efficacy of hand washing (HW) and hand washing plus rubbing with an alcohol-based solution (HWR) on numbers of total and transient flora colonies on the hands of healthcare workers (HCWs) during routine patient care was assessed. Samples were collected, using a standard bag broth technique, from the hands of 154 HCWs, before and immediately after carrying out a hand hygiene procedure. The numbers of total and transient flora colonies per plate were counted and transient pathogens were identified. A significant statistical difference between ward speciality was detected with respect to the isolation rate of transient flora. Transient hand flora were recovered from 25.3% of HCWs before carrying out the hand hygiene procedure. With respect to the disappearance and prevention of regrowth of transient flora after hand hygiene, the HWR technique was significantly more effective than HW. In conclusion, a disinfectant should be added to the hand washing process to achieve optimum protection against nosocomial infections in routine hospital practice.

Microbial flora on the hands of healthcare workers

The aims of this study were to study the process of microbial contamination of healthcare workers’ (HCWs) hands during routine patient care and simultaneously to evaluate the state of antibiotic susceptibility of microorganisms isolated from the hands of HCWs. Samples were collected from the hands of 154 HCWs working at Gaziantep University Hospital. A standard bag broth technique was employed. In addition to conventional methods, VITEK 2 automated system and API ID 32C kits were used for identification of the isolated microorganisms. Antimicrobial susceptibility of bacterial agents was performed by disc diffusion method. For antifungal susceptibility, ATB Fungus 2 kit was used. Of 154 specimens, 148 (96.1%) showed microbial growth and 160 microorganisms were isolated as single or multiple members of the resident flora. Transient flora including one or more microorganisms (n = 47) were recovered from the hands of 39 (25.3%) HCWs. All gram-positive bacteria were detected to be sensitive to vancomycin, teicoplanin and linezolid, however 30.1% of the coagulase-negative staphylococci and 40% ofStaphylococcus aureus isolates were resistant to methicillin. Multiple drug resistance was determined in Acinetobacter and Pseudomonas isolates. It can be concluded that transient flora members isolated from hands of HCWs can be causative agents of serious nosocomial infections due to their high and multiple antimicrobial resistance patterns. Key words: Healthcare workers, resident flora, transient flora, antimicrobial susceptibility.

On the evolutionary design of heterogeneous Bagging models

Bagging is a popular ensemble algorithm based on the idea of data resampling. In this paper, aiming at increasing the incurred levels of ensemble diversity, we present an evolutionary approach for optimally designing Bagging models composed of heterogeneous components. To assess its potentials, experiments with well-known learning algorithms and classification datasets are discussed whereby the accuracy, generalization and diversity levels achieved with heterogeneous Bagging are matched against those delivered by standard Bagging with homogeneous components.

Performance characteristics of a novel blood bag in‐line closure device and subsequent product quality assessment

BACKGROUNDIn high-volume processing environments, manual breakage of in-line closures can result in repetitive strain injury (RSI). Furthermore, these closures may be incorrectly opened causing shear-induced hemolysis. To overcome the variability of in-line closure use and minimize RSI, Fresenius Kabi developed a new in-line closure, the CompoFlow, with mechanical openers.STUDY DESIGN AND METHODSThe consistency of the performance of the CompoFlow closure device was assessed, as was its effect on component quality. A total of 188 RBC units using CompoFlow blood bag systems and 43 using the standard bag systems were produced using the buffy coat manufacturing method. Twenty-six CompoFlow platelet (PLT) concentrates and 10 control concentrates were prepared from pools of four buffy coats. RBCs were assessed on Days 1, 21, and 42 for cellular variables and hemolysis. PLTs were assessed on Days 1, 3, and 7 for morphology, CD62P expression, glucose, lactate, and pH. A total of 308 closures were excised after processing and the apertures were measured using digital image analysis.RESULTSThe use of the CompoFlow device significantly improved the mean extraction time with 0.46 ± 0.11 sec/mL for the CompoFlow units and 0.52 ± 0.13 sec/mL for the control units. The CompoFlow closures showed a highly reproducible aperture after opening (coefficient of variation, 15%) and the device always remained opened. PLT and RBC products showed acceptable storage variables with no differences between CompoFlow and control.CONCLUSIONSThe CompoFlow closure devices improved the level of process control and processing time of blood component production with no negative effects on product quality.

Text classification using graph mining-based feature extraction

A graph-based approach to document classification is described in this paper. The graph representation offers the advantage that it allows for a much more expressive document encoding than the more standard bag of words/phrases approach, and consequently gives an improved classification accuracy. Document sets are represented as graph sets to which a weighted graph mining algorithm is applied to extract frequent subgraphs, which are then further processed to produce feature vectors (one per document) for classification. Weighted subgraph mining is used to ensure classification effectiveness and computational efficiency; only the most significant subgraphs are extracted. The approach is validated and evaluated using several popular classification algorithms together with a real world textual data set. The results demonstrate that the approach can outperform existing text classification algorithms on some dataset. When the size of dataset increased, further processing on extracted frequent features is essential.

Evaluation of Largemouth Bass Length Limits and Catch‐and‐Release Regulations, with Emphasis on the Incorporation of Biologists' Perceptions of Largemouth Bass Length Frequency Distributions

AbstractA survey of fisheries biologists was used to assess changes in size structure of populations of largemouth bass Micropterus salmoides in relation to additional harvest regulations and to associate biologists' perceptions with various measures of length frequency changes. Biologists were presented with largemouth bass population data from regulated and control fisheries and then were asked to judge whether they perceived that the population size structure had improved over time. A blind survey approach was used (biologists did not know which populations received the additional regulations or which were control fisheries). We developed statistical models that include lake, regulation type, and biologist effects and compared them with traditional statistical analyses of regulation effectiveness. Based on perceptions of fisheries biologists about length frequency changes, 12‐in maximum length limits and catch‐and‐release regulations were about four times more likely to improve the size structure than just reliance on standard bag limits (the regulation on control fisheries). Improvement in size structure substantially beyond a statistically significant change was often needed before a regulation was considered successful by fisheries biologists. Individuals differed in their propensity to see changes as improvements. The advantages and caveats of using biologists' perceptions as a component of regulation analyses were discussed. A similar approach using angler questionnaires may yield informative data on what constitutes an improved fishery.

Out-of-bag estimation of the optimal sample size in bagging

The performance of m -out-of- n bagging with and without replacement in terms of the sampling ratio ( m / n ) is analyzed. Standard bagging uses resampling with replacement to generate bootstrap samples of equal size as the original training set m wor = n . Without-replacement methods typically use half samples m wr = n / 2 . These choices of sampling sizes are arbitrary and need not be optimal in terms of the classification performance of the ensemble. We propose to use the out-of-bag estimates of the generalization accuracy to select a near-optimal value for the sampling ratio. Ensembles of classifiers trained on independent samples whose size is such that the out-of-bag error of the ensemble is as low as possible generally improve the performance of standard bagging and can be efficiently built.

Transfusion de concentrés de globules rouges à Yaoundé, Cameroun : quelle qualité ?

As part of a quality assurance process in the transfusion service of a hospital blood bank of Yaoundé, Cameroon, a selection of units of red cell concentrates (RCC) were evaluated for volume, haemoglobin, and haematocrit levels as well as blood cell content. Blood samples were all collected into standard double blood bags containing an anticoagulant, citrate-phosphate-dextrose and adenine. During a three-month period, 35 bags intended for the preparation of the RCC were analysed. After relevant screening for transfusion transmissible infections ,and ABO and rhesus (RH1) blood grouping, the bags were centrifuged to obtain RCC. The resultant red cell bags were weighed and the volumes estimated. Full blood counts were performed on samples of the RCC using an electronic particle counter (DIANA 5, HYCEL Diagnostics, Reims, France). The results obtained showed that, based on ISO 9001: 2000 norms, there were 57, 66 and 80% of RCC respectively with volumes, hemoglobin levels as well as hematocrit that were in conformity with the norms. When the data was analysed based on the Algerian norms, 83, 66 and 95% respectively conformed. The significance of these findings and the need for establishing local norms for quality assurance in our community are discussed.

Smart Bag vs. Standard bag in the temporary substitution of the mechanical ventilation

To compare in intubated patients manually ventilated in order to mirror the ventilator, the respiratory and hemodynamic effects induced by a bag device equipped with an inspiratory gas flow-limiting valve (Smart Bag, 0-Two Medical Technologies Inc., Mississauga, ON, Canada) and a Standard bag. Non-randomized crossover study comparing 13 respiratory and eight hemodynamically paired parameters. Eight intubated patients were manually ventilated, each by three different intensive care workers yielding 24 sets of data for comparison. Data were collected during two sessions of manual ventilation, first with the Standard bag and then with the Smart Bag. Between each session, the patient was reconnected to the ventilator until return to the baseline. Patients, included after coronary surgery, were sedated and paralyzed. Intensive Care Unit, university hospital. Compared with Standard bag, the Smart Bag provided a decrease of inspiratory flow (23 +/- 4.7 vs. 47.3 +/- 16.5 l/min) with a decrease of peak pressure (13.3 +/- 2.9 vs. 21.9 +/- 7.3 cmH2O) and tidal volume (9.4 +/- 2.8 vs. 12.4 +/- 2.7 ml/kg). While the expiratory time was similar, the inspiratory time increased (1.83 +/- 0.58 vs. 1.28 +/- 0.46 s) with the Smart Bag, limiting the respiratory rate (14 +/- 5 vs. 17 +/- 6 cycles/min) and the minute volume (8.8 +/- 2.9 vs. 14.4 +/- 4.9 l/min). Finally, it limited the fall of the ETCO2 (27.9 +/- 5.1 vs. 24.3 +/- 5.7 mmHg) and probably the risks of severe respiratory alkalosis. The bags similarly affected hemodynamic states. In intubated patients manually ventilated, the Smart Bag limits the risks of excessive airway pressure and the fall of the ETCO2, with hemodynamic effects similar to those of the Standard bag.

Impaired contractility and remodeling of the upper gastrointestinal tract in diabetes mellitus type-1

To investigate that both the neuronal function of the contractile system and structural apparatus of the gastrointestinal tract are affected in patients with longstanding diabetes and auto mic neuropathy. The evoked esophageal and duodenal contractile activity to standardized bag distension was assessed using a specialized ultrasound-based probe. Twelve type-1 diabetic patients with autonomic neuropathy and severe gastrointestinal symptoms and 12 healthy controls were studied. The geometry and biomechanical parameters (strain, tension/stress, and stiffness) were assessed. The diabetic patients had increased frequency of distension-induced contractions (6.0 +/- 0.6 vs 3.3 +/- 0.5, P < 0.001). This increased reactivity was correlated with the duration of the disease (P = 0.009). Impaired coordination of the contractile activity in diabetic patients was demonstrated as imbalance between the time required to evoke the first contraction at the distension site and proximal to it (1.5 +/- 0.6 vs 0.5 +/- 0.1, P = 0.03). The esophageal wall and especially the mucosa-submucosa layer had increased thickness in the patients (P < 0.001), and the longitudinal and radial compressive stretch was less in diabetics (P < 0.001). The esophageal and duodenal wall stiffness and circumferential deformation induced by the distensions were not affected in the patients (all P > 0.14). The impaired contractile activity with an imbalance in the distension-induced contractions likely reflects neuronal abnormalities due to autonomic neuropathy. However, structural changes and remodeling of the gastrointestinal tract are also evident and may add to the neuronal changes. This may contribute to the pathophysiology of diabetic gut dysfunction and impact on future management of diabetic patients with gastrointestinal symptoms.

Retention of ventilation skills of emergency nurses after training with the SMART BAG ® compared to a standard bag–valve–mask

Studies show that nurses retain resuscitation skills poorly and that retention of ventilation skills is particularly difficult. We formed the hypothesis that the SMART BAG (SB, O-Two Medical Technologies Inc., Canada), i.e. a bag-valve-mask device with a pressure/flow responsive valve, would assist nurses in providing more efficient ventilation six months after training. Prior to training, 39 emergency nurses performed CPR for 2 min, in pairs, using a standard bag-valve-mask device (STBVM, Laerdal, Norway) to assess their base line skills. A CPR training manikin (Simulaids, USA) equipped with a PEEP valve in the oesophagus set at 20 cm H2O was used to measure inspiratory time, tidal volume (Vt), peak pressure and gastric insufflation (GI). Immediately following training they were tested using an O-Two STBVM and a SB. Half of the nurses were retested after three months, the other half after six months. Efficient ventilation was defined as a mean Vt>400 ml and GI<50 ml in 1 min. Before training, only 16% of nurses ventilated efficiently: 63% had GI and 28% had Vt<400 ml. Three months after training the efficiency of the STBVM and the SB was high (81 and 75%, respectively). Six months after training, there was a trend towards higher efficiency for the SB (63%) compared to the STBVM (25%) (p=0.07). For instances with the STBVM producing a Vt>400 ml, those without GI had a lower peak pressure than those with GI (7.8 cm H2O versus 17.7 cm H2O, p=0.0001) and showed a trend towards a longer inspiratory time (1.28 s versus 1.08 s, p=0.08). Of all efficient ventilations with a STBVM, 26% had a Vt>600 ml. Six months after training, nurses ventilated at least as efficiently with the SB, compared with the STBVM. This illustrates the ability of the SB to compensate for the deterioration over time in skill. On the other hand, training with a STBVM should focus primarily on prolonging the inspiratory time, and therefore the peak pressure, whilst maintaining an adequate Vt.

Lattice QCD constraints on hybrid and quark stars

A QCD-motivated dynamical-quasiparticle model with parameters adjusted to reproduce the lattice-QCD equation of state is extrapolated from the region of high temperatures and moderate baryonic densities to the domain of high baryonic densities and zero temperature. The resulting equation of state matched with realistic hadronic equations of state predicts a phase transition into the quark phase at higher densities than those reachable in neutron star interiors. This excludes the possibility of the existence of hybrid (hadron-quark) stars. Pure quark stars are possible and have low masses, small radii, and very high central densities. Similar results are obtained for a simple bag model with massive quarks, fitted to reproduce the same lattice results. Self-bound quark matter is also excluded within these models. Uncertainties in the present extrapolation are discussed. Comparison with standard bag models is made.

高酸素透過性バッグによる高単位血小板の室温長期（９日間）保存

The performance of platelet concentrates (PCs) for transfusion is greatly limited by their short shelf life. A highly oxygen-permeable container (PO-80, Kawasumi, Tokyo) was developed as an approach to extending the shelf-life of PCs. Here, we compared PCs containing a high number of platelets (4×1011/250mL plasma/bag) stored in a PO-80 container with those stored in a standard bag (control) (PL2410, Baxter Healthcare, USA) by monitoring platelet biochemical and functional parameters for up to 9 days.Three of six bags in the control bags had a pH below 6.2 by day 7. These PCs also showed the disappearance of swirling, a rapid drop in glucose combined with a marked increase in lactic acid, accelerated P-selectin expression on platelets and an earlier decrease in both %HSR and aggregation of platelets in the observation period. In contrast, only one of six bags had a pH below 6.2 by day 9 in the PO-80 container. Further, there was no disappearance of swirling. Other biochemical and functional parameters of PCs in PO-80 showed a slower change compared with those in the control bag. Overall, the in vitro characteristics of PCs showed a lower level of deterioration over 9 days when stored in a higher oxygen-permeable container than in a conventional bag. It is suggested that this newly developed highly oxygen-permeable PO-80 container will preserve adequate platelet function for longer storage periods and with higher PC yields.