A privacy-preserving distributed transfer learning in activity recognition

Telecommunication Systems - Along with the widespread use of smartphones, activity recognition using embedded inertial sensors has intrigued researchers. The learning and employing activity...     

Application of meta-analysis in plant pathology: a case study examining the impact of fungicides on wheat yield loss from the yellow spot—septoria nodorum blotch disease complex in Western Australia

This work was done to demonstrate the opportunities provided by application of meta-analysis in plant pathology. It was a case study used to determine the effectiveness of foliar fungicides in minimising yield loss from a complex of yellow spot (Pyrenophora tritici-repentis) and septoria nodorum blotch (Stagonospora nodorum, teleomorph: Phaeosphaeria nodorum) (YS-SNB disease complex) on wheat in the northern grain-belt of Western Australia. Forty-seven datasets of experimental results from 14 growing seasons, using 18 varieties sprayed one to three times, predominantly with tebuconazole or propiconazole fungicides, were analysed. Across the datasets, the wheat yield gain from fungicide application was 297 kg ha^?1 with a 95 % confidence interval of 11.6 kg ha^?1. Significant yield gains resulted from single or multiple applications of fungicides. Both propiconazole and tebuconazole, increased yield of wheat affected by the YS–SNB disease complex, with yield gain from propiconazole being greater than that from tebuconazole. Yield response varied significantly among crop growing seasons. Meta-analysis was able to aggregate a large number of experimental results and answer important questions related to the variables that influenced those results; in this case the effectiveness of fungicides in minimising yield loss from the YS-SNB disease complex on wheat. It also identified areas where further research needs to be done. It is concluded that meta-analysis has the potential to contribute to similar analyses in other crop disease systems.     

Uncertainty Analysis of a Two-dimensional Hydrodynamic Model

First-Order Second Moment (FOSM) and Monte Carlo analysis were applied to characterize the uncertainty in selected water levels and velocities simulated by a two-dimensional hydrodynamic model of the Upper St. Lawrence River downstream from Lake Ontario. The analysis utilized an application of the Resource Management Associates’ RMA2 model. Both FOSM and Monte Carlo analysis provided similar estimates of uncertainty, with Monte Carlo analysis results being 15% less than FOSM. Based on the findings of this work, the FOSM is preferred. FOSM provides a conservative estimate of the uncertainty and it is simpler to apply than Monte Carlo analysis, requiring less information and fewer model executions. FOSM also provides an immediate indication of the primary contributors to the uncertainty in the output, where Monte Carlo analysis requires additional effort to do the same. Results indicate that the parameter describing bottom resistance using Manning's n contributed more to model uncertainty than other factors investigated. The uncertainty in and sensitivity in Manning's n is large which results in a significant amount of uncertainty in the model outputs is contributed by this parameter. The calculations described in this study show that uncertainty analysis is a practical addition to the two-dimensional hydrodynamic modelling process. It provides insight to the model developer, quantifying how good the model actually is. It also provides a measure of the accuracy of the model for future model developers or clients using hydrodynamic modeling outputs     

Location allocation modeling for healthcare facility planning in Malaysia

Malaysia has seen tremendous growth in the standard of living and household per capita income. The demand for a more systematic and efficient planning has become increasingly more important, one of the keys to achieving a high standard in healthcare. In this paper, a Maximal Covering Location Problem (MCLP) is used to study the healthcare facilities of one of the districts in Malaysia. We address the limited capacity of the facilities and the problem is formulated as Capacitated MCLP (CMCLP). We propose a new solution approach based on genetic algorithm to examine the percentage of coverage of the existing facilities within the allowable distance specified/targeted by Malaysian government. The algorithm was shown to generate good results when compared to results obtained using CPLEX version 12.2 on a medium size problem consisting of 179 nodes network. The algorithm was extended to solve larger network consisting of 809 nodes where CPLEX failed to produce non-trivial solutions. We show that the proposed solution approach produces significant results in determining good locations for the facility such that the population coverage is maximized.     

优化水泥厂能源管理和生产运作

Lehigh水泥公司在对其位于马里兰州Union Bridge的工厂进行现代化管理的过程中，集成运用了过程控制和能源管理以达到改善整个工厂的生产效率的目的。     

Designing human–machine interfaces for naturalistic perceptions,decisions and actions occurring in emergency situations

Humans experience emotional arousal, when threat to life and limb is imminent, such as in high stakes, emergency situations that occur in sociotechnical systems like fire fighting, aviation, combat, etc. Emotional arousal has both advantages and disadvantages. On the downside, it introduces perceptual distortions and biases and inability to process symbolic information and alters motor abilities. As emergency human–machine interfaces (HMIs) will be used to bring a system under control, i.e. to avoid or minimise losses, they should be designed to accommodate the human capacities that have been altered by danger-induced emotional arousal. A model called direct perception–action coupling (DPAC) that combines direct perception and embodied cognition, which is not predicated on higher cognitive functions, has been developed to describe the available capacities of an operator under stress. Finally, the conceptual application of human performance theories under stress, through the DPAC model, to design emergency HMIs is discussed with examples.     

Erythrocytes from healthy smokers bind more bilirubin than the erythrocytes from healthy non-smokers

Cigarette smoking is an adverse prognostic factor for health. Its damaging effects on many enzymatic and cellular activities are well known. The present study was carried out to evaluate whether there is a difference in the binding of bilirubin to the erythrocytes from healthy smokers and non-smokers. The results suggest that the binding of bilirubin to the erythrocytes from healthy smokers as well as in vitro smoked erythrocytes is significantly higher than that of healthy non-smokers.