Rival-Model Penalized Self-Organizing Map

As a typical data visualization technique, self-organizing map (SOM) has been extensively applied to data clustering, image analysis, dimension reduction, and so forth. In a conventional adaptive SOM, it needs to choose an appropriate learning rate whose value is monotonically reduced over time to ensure the convergence of the map, meanwhile being kept large enough so that the map is able to gradually learn the data topology. Otherwise, the SOM's performance may seriously deteriorate. In general, it is nontrivial to choose an appropriate monotonically decreasing function for such a learning rate. In this letter, we therefore propose a novel rival-model penalized self-organizing map (RPSOM) learning algorithm that, for each input, adaptively chooses several rivals of the best-matching unit (BMU) and penalizes their associated models, i.e., those parametric real vectors with the same dimension as the input vectors, a little far away from the input. Compared to the existing methods, this RPSOM utilizes a constant learning rate to circumvent the awkward selection of a monotonically decreased function for the learning rate, but still reaches a robust result. The numerical experiments have shown the efficacy of our algorithm     

Optical appearance of alpha particle tracks in CR-39 SSNTD

A ray tracing method based on geometrical optics was used to study the tracks from alpha particles with different energies and with an incident angle of 50°. The transmission operation mode of the microscope is simulated. Considering the distribution of light intensities from the tracks, the mean and the 80% percentile gray levels from real experiments are proposed as quantitative variables to differentiate among tracks. The gray level properties for the same track for different exposures can vary to large extents. We introduce three variables, κ, λ and ε, to make empirical corrections. It is interesting to see that these coefficients are very consistent for the same alpha particle track despite the very different gray level properties. Gray level results have been obtained for tracks from alpha particles with 50° incident angle and different incident energies. However, the track depths cannot be predicted by any one of the coefficients κ, λ and ε. Multivariate analyses can help separate the tracks corresponding to different alpha energies. By using discriminant analysis with κ, λ and ε as independents, effectively all alpha energies can be determined with an accuracy of ±0.5 MeV.     

Practice characteristics that lead to 21st century learning outcomes

Abstract The key research question for this study was to ask whether or not innovative teaching practices would lead to the development of learning outcomes essential for preparing the younger generation for the challenges of life in the knowledge society of the 21st century, and if so, how are the pedagogical features related to the different learning outcomes. Preliminary analyses of the case study data collected from the SITES M2 Study in Hong Kong reveal that where the development of more significant learning gains were observed, the cases possess characteristics additional to the criteria defined in the Study for selection of innovation. More importantly, it was found that the impact of the pedagogical practices was not determined simply by the aggregation of characteristics of the practices per se , nor on the technologies used, but on whether 'empowerment' permeates the curriculum goal and process. Further, this paper claims that these affective and socio-cognitive learning outcomes are more important as preparation for lifelong learning in the 21st century than 'knowledge management competencies'.     

EIR1, a root-specific protein involved in auxin transport, is required for gravitropism in Arabidopsis thaliana

The EIR1 gene of Arabidopsis is a member of a family of plant genes with similarities to bacterial membrane transporters. This gene is expressed only in the root, which is consistent with the phenotypes of the eir1 mutants-the roots are agravitropic and have a reduced sensitivity to ethylene. The roots of eir1 mutants are also insensitive to the excess auxin produced by alf1-1 and fail to induce an auxin-inducible gene in the expansion zone. Although they fail to respond to internally generated auxin, they respond normally to externally applied auxin. Expression of the EIR1 gene in Saccharomyces cerevisiae confers resistance to fluorinated indolic compounds. Taken together, these data suggest that the EIR1 protein has a root-specific role in the transport of auxin.     

Native valve infective endocarditis in elderly and younger adult patients: comparison of clinical features and outcomes with use of the Duke criteria and the Duke Endocarditis Database

The effect of age on the presentation and outcome of infective endocarditis (IE) is unclear. Many of the available data are based on analyses of mixed populations of patients including intravenous drug users or those with prosthetic valve endocarditis or native valve IE. We used the Duke criteria to compare the characteristics of 44 episodes of definite native valve IE in elderly patients (> 64 years old) with the characteristics of 64 similarly defined episodes of native valve IE in younger, nonintravenous-drug-using adult patients (> 29 years and < 60 years old). Our data suggest that the clinical presentation, characteristics, and outcome of native valve IE are similar for elderly patients and younger adult patients, although elderly patients were hospitalized an average of 12 days longer. Although we found that the occurrence of renal failure and cerebral embolism during an episode of IE was associated with higher rates of death (odds ratios, 4.8 and 4.0, respectively), age was not a significant contributor to mortality.     

Utility of ischemic scores in the differential diagnosis of Alzheimer's disease and ischemic vascular dementia

Despite new developments in the concept of vascular dementia, the Hachinski Ischemic Score (HIS) and its modified versions continue to be widely used in the clinical differentiation of Alzheimer's disease (AD) and ischemic vascular dementia (IVD). The sensitivity of the HIS and two modified versions in the diagnosis of AD, IVD, and single infarcts in a large, geriatric population with mild cognitive impairment (N = 100) was evaluated. Sensitivity for identification of AD was greater than 90% but was less than 70% for IVD. Over one third of patients with one or more infarcts on computed tomographic brain scans and 63% of mixed cases were classified as having probable AD. It is concluded that ischemic scores may be useful at predicting prevalence rates if individual case accuracy is ignored. Despite being sensitive to identifying AD, ischemic scores are insensitive to both cerebral infarction and IVD and cannot reliably exclude IVD. Finally, patients with mixed dementia should not be expected to have intermediate scores.     

Effect of pulmonary artery stenoses on the cardiopulmonary response to exercise following repair of tetralogy of Fallot

Data from exercise tests, echocardiograms, and lung perfusion scans were analyzed to determine whether the excessive minute ventilation (VE) often encountered among patients with tetralogy of Fallot is due to ventilation-perfusion mismatch secondary to branch pulmonary artery stenoses. Patients with branch PA stenoses had lower peak oxygen consumptions and higher VE during exercise than did patients without stenoses, and a strong correlation existed between the degree of pulmonary blood flow maldistribution on lung perfusion scan and the amount of excessive VE during exercise.