Variable Reduction Technique Research Articles

Medical error reporting, culture, and infrastructure.

245 Background: Much work has focused on the importance of culture as a key factor in advancing the patient safety discussion. No assumption can be made that a culture of safety equates to embracing responsibility to report when errors and near misses occur. Underreporting exists. Without data, the opportunity to redesign care and address system failures is diminished. This research aimed to answer: what impact do various factors have on influencing the likelihood of reporting across a hospital network of oncology providers? Methods: The AHRQ Hospital Culture on Safety was used coupled with 15 custom questions of interest. Dependent variables (DVs) included: structure for reporting (definition, policy and ease), incentives to report, supervisor/employee relationship and management support. Independent variables (IVs) included number of errors reported and perceived frequency by which events are reported, influenced by tenure and extent of patient interaction. Higher scores associated with the IVs were posited to be positively associated with the errors reported as well as the reporting frequency. A principle component analysis was conducted as a variable reduction technique to affirm factors; the reliability of factors tested; and multiple regression analyses conducted to test the strength of relationships. Results: 2,103 surveys were completed. The 4 factors are found to weakly predict the DVs. Infrastructure had the greatest association in both models. Tenure and patient interaction influenced the likelihood of reporting. Conclusions: This research moves beyond culture in ascertaining factors pivotal in the underreporting of errors and near misses. An infrastructure promoting reporting must also exist ensuring staff know what, how, when, and why to report, linking data to action.[Table: see text]

Uncertainty estimation and misclassification probability for classification models based on discriminant analysis and support vector machines

Uncertainty estimation provides a quantitative value of the predictive performance of a classification model based on its misclassification probability. Low misclassification probabilities are associated with a low degree of uncertainty, indicating high trustworthiness; while high misclassification probabilities are associated with a high degree of uncertainty, indicating a high susceptibility to generate incorrect classification. Herein, misclassification probability estimations based on uncertainty estimation by bootstrap were developed for classification models using discriminant analysis [linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA)] and support vector machines (SVM). Principal component analysis (PCA) was used as variable reduction technique prior classification. Four spectral datasets were tested (1 simulated and 3 real applications) for binary and ternary classifications. Models with lower misclassification probabilities were more stable when the spectra were perturbed with white Gaussian noise, indicating better robustness. Thus, misclassification probability can be used as an additional figure of merit to assess model robustness, providing a reliable metric to evaluate the predictive performance of a classifier.

The matching relaxation for a class of generalized set partitioning problems

This paper introduces a discrete relaxation for the class of combinatorial optimization problems which can be described by a set partitioning formulation under packing constraints. We present two combinatorial relaxations based on computing maximum weighted matchings in suitable graphs. Besides providing dual bounds, the relaxations are also used on a variable reduction technique and a matheuristic. We show how that general method can be tailored to sample applications, and also perform a successful computational evaluation with benchmark instances of a problem in maritime logistics.

Novel optical efficiency formulas for parabolic trough solar collectors: Computing method and applications

This paper presents a novel computing method of optical efficiency fitting formulas and its applications for parabolic trough solar collectors, by combining the Monte Carlo ray-tracing methodology with the population-based particle swarm optimization algorithm. The expressions of the optical efficiency fitting formulas are deduced by means of analyzing the Monte Carlo ray-tracing data samples with a variable reduction technique, while the fitting formula coefficients are obtained using the particle swarm optimization algorithm. The computing time of the annual optical efficiency reduces to less than 10−7 s, from about a day as previously calculated by a once powerful Monte Carlo ray-tracing model. Thus, it makes the once time-consuming long-time average optical efficiency calculations or corresponding population-based optimizations much more feasible and efficient. A dynamic particle swarm size factor is also proposed to improve the particle swarm optimization performance, which reduces nearly 90% of the particle swarm optimization computing time. Numerical results calculated by the Monte Carlo ray-tracing model and the obtained optical efficiency fitting formulas were compared with the corresponding reference data. Good agreements were obtained, proving that the computing method and model are reliable. The obtained optical efficiency fitting formula of parabolic trough solar collectors is then applied successfully to efficiently study the characteristics of long-time average optical efficiencies under different geometrical configurations, geographical locations, tracking models and corresponding optical or geometrical optimizations. This proposed novel method offers a useful option of high potential for further applications on efficient optical efficiency related analyses and corresponding population-based optimizations for parabolic trough solar collectors.

Training Deep Convolutional Neural Networks with Resistive Cross-Point Devices

In a previous work we have detailed the requirements for obtaining maximal deep learning performance benefit by implementing fully connected deep neural networks (DNN) in the form of arrays of resistive devices. Here we extend the concept of Resistive Processing Unit (RPU) devices to convolutional neural networks (CNNs). We show how to map the convolutional layers to fully connected RPU arrays such that the parallelism of the hardware can be fully utilized in all three cycles of the backpropagation algorithm. We find that the noise and bound limitations imposed by the analog nature of the computations performed on the arrays significantly affect the training accuracy of the CNNs. Noise and bound management techniques are presented that mitigate these problems without introducing any additional complexity in the analog circuits and that can be addressed by the digital circuits. In addition, we discuss digitally programmable update management and device variability reduction techniques that can be used selectively for some of the layers in a CNN. We show that a combination of all those techniques enables a successful application of the RPU concept for training CNNs. The techniques discussed here are more general and can be applied beyond CNN architectures and therefore enables applicability of the RPU approach to a large class of neural network architectures.

Built Environment, Travel Attitudes and Travel Behaviour: Quasi-Longitudinal Analysis of Links in the Case of Greeks Relocating from US to Greece

In this study, the possible causal links between the built environment, travel attitudes and travel behaviour of people that have moved between totally different urban and transportation contexts were investigated. A quasi-longitudinal design was developed to collect data about the perceived neighbourhood characteristics, neighbourhood preferences, travel attitudes and changes in car, walking and bicycle usage of 51 Greeks who relocated from the US to Greece. Variable reduction techniques were applied prior to developing our models because of the small sample size. The results of this study offer support for a causal relationship between the built environment and travel behaviour identified in the existing literature. For instance, longer travel time to city center was associated with higher car use, higher density neighbourhoods and increased accessibility to local amenities were associated with increased bike use and better access to a district shopping center was associated with more walking after relocation. Our results also showed that in contexts like Greece, where transport infrastructures are not adequately developed, lack of safe bike conditions and easy access to public transportation are important determinants of bicycle use and walking. Thus, according to our results, promoting sustainable mobility in contexts like Greece would require not only enhancing accessibility through relocation of activities, but also by improving infrastructures for public transport, bicycle and walking.

A methodology for relating wetland configuration to human disturbance in Alberta

Widespread loss of wetland ecosystems resulting from human land use highlights the need for a reclamation strategy that can sustain wetland ecosystem services. Since wetland function partly depends on landscape structure, reclamation and monitoring can be aided by knowing the differences in wetland configuration between undisturbed and disturbed landscapes. Identify a parsimonious set of landscape metrics for quantifying wetland configuration and land-cover composition, and quantify how these metrics vary with anthropogenic disturbance. A large suite of landscape metrics quantifying area/edge, shape, aggregation, and diversity were calculated for landscapes in the Grassland, Parkland, and Boreal Natural Regions of Alberta. Variable reduction techniques were applied to identify representative metrics. These representative metrics were related to anthropogenic disturbance using non-parametric tests. The spatial configuration of wetlands in low-disturbance and high-disturbance landscapes were significantly different from other landscapes. Aggregation metrics were the most commonly identified measures of wetland configuration independent of wetland-proportion in the landscape. Our findings provide insight for reclamation and monitoring by showing that some aspects of wetland configuration vary independently of composition, and therefore both need to be considered when parameterizing the design of reclaimed landscapes. We suggest that using landscape metrics in a reference condition approach is appropriate for evaluating landscape degradation and for setting landscape reclamation targets and monitoring programs.

Rice quality profiling to classify germplasm in breeding programs

The objective of this work was to define a quality space for assessing rice varieties. Eleven long grain varieties, seven commercial and four new advanced lines were assessed to obtain complete quality profile considering appearance, physicochemical parameters, water absorption behaviour, pasting profile and textural attributes. Commercial varieties were chosen to provide the widest variation in properties, applying the variability analysis concepts of the Taguchi method, including Japonicas, Indicas, hybrids and aromatics. Quality parameters were measured in five different dimensions of quality space (totalling 50). Variable reduction techniques were applied to chose 3 parameters in each dimension (totalling 15 quality indicators) that would describe the whole space with greatest orthogonality, accuracy and yet explaining a significant proportion of the whole variance of data. The analysis of the quality space thus defined and similarities between varieties is illustrated with the conclusion of how the 4 new advanced lines perform in terms of quality behaviour, where it is concluded that one of them is very promising as an improvement over European (Indica) towards the behaviour of a pure Guyana (Indica), whereas 3 others have significant shortcomings in various aspects of the quality space compared to all others, albeit their greater closeness to Japonicas.

Vocal Acoustic Analysis – Classification of Dysphonic Voices with Artificial Neural Networks

Voice acoustic analysis is becoming nowadays a useful tool for detection of laryngological pathologies. This techniques enables a non-invasive and low cost assessment of voice disorders allowing a more efficient fast and objective diagnosis, permitting the patients to get a suitable treatment. In this work, the best predictors/parameters for diagnose of dysphonia were experimented. A vector made up of 4 Jitter parameters, 4 Shimmer parameters and Harmonic to Noise Ratio (HNR), determined from 3 different vowels at 3 different tones, in a total of 81 features, was used. Variable selection and dimension reduction techniques such as hierarchical clustering, multilinear regression analysis and principal component analysis (PCA) was applied. For the classification models based on artificial neural network (ANN) was used. The methods/models found allowed us to obtain an Accuracy of 100% for female voices and 90% for male voices using only Jitter Shimmer and HNR parameters.

Combinatorial Relaxation Bounds and Preprocessing for Berth Allocation Problems

We investigate an optimization problem in container ports, for which previous models based on generalized set partitioning formulations have been studied. We describe two combinatorial relaxations based on computing maximum weighted matchings in suitable graphs, providing dual bounds and a variable reduction technique.

Efficiency factors in OECD banks: A ten-year analysis

This paper presents a performance assessment of 128 banks from 23 OECD countries from 2004 to 2013, using different financial criteria that emulate the CAMELS rating system. A robust TOPSIS approach for assessing bank efficiency is also developed and presented. First, alternative variable reduction techniques are employed to extract the major factors within each CAMELS criterion. This is done to mitigate collinearity issues. Then, TOPSIS is used to measure bank performance based upon these factors, equally weighted. A comprehensive analysis based on a weighted linear optimization model for multi-criteria classification is also performed, which detects any discrepancies from the original scores. Lastly, censored quantile regressions are combined with bootstrapped TOPSIS scores to produce a model for predicting the impact of different contextual variables on different efficiency quantiles. Results reveal that the effects of ownership, trend, and origin of the bank may vary with respect to efficiency levels, whether high or low.

Integrated analysis of multidimensional omics data on cutaneous melanoma prognosis

Multiple types of genetic, epigenetic, and genomic changes have been implicated in cutaneous melanoma prognosis. Many of the existing studies are limited in analyzing a single type of omics measurement and cannot comprehensively describe the biological processes underlying prognosis. As a result, the obtained prognostic models may be less satisfactory, and the identified prognostic markers may be less informative. The recently collected TCGA (The Cancer Genome Atlas) data have a high quality and comprehensive omics measurements, making it possible to more comprehensively and more accurately model prognosis. In this study, we first describe the statistical approaches that can integrate multiple types of omics measurements with the assistance of variable selection and dimension reduction techniques. Data analysis suggests that, for cutaneous melanoma, integrating multiple types of measurements leads to prognostic models with an improved prediction performance. Informative individual markers and pathways are identified, which can provide valuable insights into melanoma prognosis.

Universally Conserved Relationships between Nuclear Shape and Cytoplasmic Mechanical Properties in Human Stem Cells.

The ability of cells to proliferate, differentiate, transduce extracellular signals and assemble tissues involves structural connections between nucleus and cytoskeleton. Yet, how the mechanics of these connections vary inside stem cells is not fully understood. To address those questions, we combined two-dimensional particle-tracking microrheology and morphological measures using variable reduction techniques to measure whether cytoplasmic mechanics allow for discrimination between different human adherent stem cell types and across different culture conditions. Here we show that nuclear shape is a quantifiable discriminant of mechanical properties in the perinuclear cytoskeleton (pnCSK) of various stem cell types. Also, we find the pnCSK is a region with different mechanical properties than elsewhere in the cytoskeleton, with heterogeneously distributed locations exhibiting subdiffusive features, and which obeys physical relations conserved among various stem cell types. Finally, we offer a prospective basis to discriminate between stem cell types by coupling perinuclear mechanical properties to nuclear shape.

On a Practical, Integer-Linear Programming Model for Delete-Free Tasks and its Use as a Heuristic for Cost-Optimal Planning

We propose a new integer-linear programming model for the delete relaxation in cost-optimal planning. While a straightforward IP for the delete relaxation is impractical, our enhanced model incorporates variable reduction techniques based on landmarks, relevance-based constraints, dominated action elimination, immediate action application, and inverse action constraints, resulting in an IP that can be used to directly solve delete-free planning problems. We show that our IP model is competitive with previous state-of-the-art solvers for delete-free problems. The LP-relaxation of the IP model is often a very good approximation to the IP, providing an approach to approximating the optimal value of the delete-free task that is complementary to the well-known LM-cut heuristic. We also show that constraints that partially consider delete effects can be added to our IP/LP models. We embed the new IP/LP models into a forward-search based planner, and show that the performance of the resulting planner on standard IPC benchmarks is comparable with the state-of-the-art for cost-optimal planning.

Output-Constraint Handling and Parallelization for Oil-Reservoir Control Optimization by Means of Multiple Shooting

Summary We propose to formulate and solve the reservoir-control optimization problem with the direct multiple-shooting method. This method divides the optimal-control problem prediction horizon in smaller intervals that one can evaluate in parallel. Further, output constraints are easily established on each interval boundary and as such hardly affect computation time. This opens new opportunities to include state constraints on a much broader scale than is common in reservoir optimization today. However, multiple shooting deals with a large number of variables because it decides on the boundary-state variables of each interval. Therefore, we exploit the structure of the reservoir simulator to conceive a variable-reduction technique to solve the optimization problem with a reduced sequential quadratic-programming algorithm. We discuss the optimization-algorithm building blocks and focus on structure exploitation and parallelization opportunities. To demonstrate the method's capabilities to handle output constraints, the optimization algorithm is interfaced to an open-source reservoir simulator. Then, on the basis of a widely used reservoir model, we evaluate performance, especially related to output constraints. The performance of the proposed method is qualitatively compared with a conventional method.

Integrated Berth Allocation and Quay Crane Assignment Problem: Set partitioning models and computational results

Most of the operational problems in container terminals are strongly interconnected. In this paper, we study the integrated Berth Allocation and Quay Crane Assignment Problem in seaport container terminals. We will extend the current state-of-the-art by proposing novel set partitioning models. To improve the performance of the set partitioning formulations, a number of variable reduction techniques are proposed. Furthermore, we analyze the effects of different discretization schemes and the impact of using a time-variant/invariant quay crane allocation policy. Computational experiments show that the proposed models significantly improve the benchmark solutions of the current state-of-art optimal approaches.

Health-related and psychosocial concerns about transplantation among patients initiating dialysis.

Disparities in kidney transplantation remain; one mechanism for disparities in access to transplantation (ATT) may be patient-perceived concerns about pursuing transplantation. This study sought to characterize prevalence of patient-perceived concerns, explore interrelationships between concerns, determine patient characteristics associated with concerns, and assess the effect of concerns on ATT. Prevalences of 12 patient-perceived concerns about pursuing transplantation were determined among 348 adults who recently initiated dialysis, recruited from 26 free-standing dialysis centers around Baltimore, Maryland (January 2009-March 2012). Using variable reduction techniques, concerns were clustered into two categories (health-related and psychosocial) and quantified with scale scores. Associations between patient characteristics and concerns were estimated using modified Poisson regression. Associations between concerns and ATT were estimated using Cox models. The most frequently cited patient-perceived concerns were that participants felt they were doing fine on dialysis (68.4%) and felt uncomfortable asking someone to donate a kidney (29.9%). Older age was independently associated with having high health-related (adjusted relative risk, 1.35 [95% confidence interval, 1.20 to 1.51], for every 5 years older for those ≥ 60 years) or psychosocial (1.15 [1.00 to 1.31], for every 5 years older for those aged ≥ 60 years) concerns, as was being a woman (1.72 [1.21 to 2.43] and 1.55 [1.09 to 2.20]), having less education (1.59 [1.08 to 2.35] and 1.77 [1.17 to 2.68], comparing postsecondary education to grade school or less), and having more comorbidities (1.18 [1.08 to 1.30] and 1.18 [1.07 to 1.29], per one comorbidity increase). Having never seen a nephrologist before dialysis initiation was associated with high psychosocial concerns (1.48 [1.01 to 2.18]). Those with high health-related (0.37 [0.16 to 0.87]) or psychosocial (0.47 [0.23 to 0.95]) concerns were less likely to achieve ATT (median follow-up time 2.2 years; interquartile range, 1.6-3.2). Patient-perceived concerns about pursuing kidney transplantation are highly prevalent, particularly among older adults and women. Reducing these concerns may help decrease disparities in ATT.

AN EFFICIENT AND VARIABLE INTERVAL REKEYING FOR DYNAMIC AD HOC MULTICAST GROUP USING KEY PATH REDUCTION

Key distribution is one of the major issues in secure ad hoc multicast group communication. There has been an extensive research on rekeying, to reduce cost. In this study, we propose an efficient and scalable batch rekeying for dynamic ad hoc multicast group, with variable interval and key path reduction techniques. The proposed scheme overcomes the major existing issues like inefficiency in using the keys that are generated and distributed, the sync issue in which a user tries to decrypt a data using an irrelevant key, imbalance in network traffic and latency in the key server response to the user request, for leaving and joining at once. Generally, a central key server is used to govern all the above issues. The proposed scheme excludes the usage of a central key server by generating the group key in the individual nodes, which minimizes the communication overhead and the number of keys that each user possesses. The proposed scheme also reduces the depth of the tree very effectively, when a user joins or leaves the group, using key path reduction technique and ensures forward and backward secrecy. The simulation result shows better performance when compared to individual, regular batch and periodic rekeying.

Surgical Management of Pediatric Radial Neck Fractures

Management of pediatric radial neck fractures is controversial regarding acceptable alignment, variable reduction techniques, and suboptimal outcomes. The purpose of this study was to assess the characteristics, management, and results in a surgical cohort, in efforts to identify prognostic factors and offer treatment suggestions. It was hypothesized that less invasive reduction maneuvers would precede open reduction and that worse results would correlate with fracture severity, open reduction, and the presence of associated injuries. Retrospective analysis of 151 children in whom a radial neck fracture had been surgically treated from 2001 to 2011 was performed. The mean age (and standard deviation) and duration of follow-up were 8.4 ± 2.9 years and 13.3 ± 20.0 months, respectively; 40% of the patients were male. A successful clinical result was defined as elbow flexion of ≥120°, flexion contracture of <20°, forearm rotation of ≥90° with ≥45° of supination and pronation, and no complications. An isolated radial neck fracture occurred in 54% of the children. The mean angulation and displacement improved from 43° ± 19° and 37% ± 35%, respectively, before treatment to 13° ± 7° and 0.9% ± 4% after treatment (p < 0.001). Twenty-two procedural combinations were used to treat these patients, and 67% of the open reductions were not preceded by percutaneous or closed reduction attempts. Among 131 patients with adequate follow-up, 31% had an unsuccessful outcome. An age of ten years or more (odds ratio [OR] = 5.85, p = 0.001), a time to surgery of two days or less (OR = 4.73, p = 0.02), and greater fracture displacement (OR = 1.25 per 10%, p = 0.001) were independent predictors of unsuccessful outcomes. Increased fracture severity and open reduction were associated with poor results, although the presence of concomitant injuries was not. It is predicted that closed manipulation will fail for half of fractures angulated ≥36°, and that half of fractures displaced ≥65% will require open reduction. The predicted frequency of unsuccessful outcomes is 50% with 76% displacement. There continues to be great variation in the approach to treatment of displaced radial neck fractures in children. Suboptimal results occurred in 31% of the patients in this series, with worse results in patients older than ten years, who had increased fracture severity, and who underwent open reduction. Less invasive reduction methods should precede open reduction whenever possible.

A Collective Study of PCA and Neural Network based on COCOMO for Software Cost Estimation

Estimating cost is a very wearisome activity in all aspect. A person with broad scope and good thinking for the future makes more precise decisions. It helps in governing and planning the software risks which are admirably correct and precise. In 1960 regression analysis and mathematical formulae were practiced to determine cost. We need to think more than simply putting numbers into a formula and accept the results to attaining the accuracy of software cost estimation. The changing methods of estimating software cost have made the researchers to think diversely. Barry Bohem birthed COCOMO model for software cost estimation in 1981 which is considered to be more efficient as compared to previous models. Thereafter number of researchers has been trying to improve the efficiency by keeping the base of COCOMO model. The paper drafts a novel variable reduction technique called feed-forward neural network with PCA to measure the estimation model accuracy. This is based on a COCOMO sample data set which collects and maintains a large software project data repository. PCA is a kind of classification method which can reduces number of factors into a few absolute factors.