Search

This article presents a study based on evaluating different techniques to automatically recognize the basic emotions of people with Down syndrome, such as anger, happiness, sadness, surprise, and neutrality, as well as the statistical analysis of the Facial Action Coding System, determine the symmetry of the Action Units present in each emotion, identify the facial features that represent this group of people. First, a dataset of images of faces of people with Down syndrome classified according to their emotions is built. Then, the characteristics of facial micro-expressions (Action Units) present in the feelings of the target group through statistical analysis are evaluated. This analysis uses the intensity values of the most representative exclusive action units to classify people's emotions. Subsequently, the collected dataset was evaluated using machine learning and deep learning techniques to recognize emotions. In the beginning, different supervised learning techniques were used, with the Support Vector Machine technique obtaining the best precision with a value of 66.20%. In the case of deep learning methods, the mini-Xception convolutional neural network was used to recognize people's emotions with typical development, obtaining an accuracy of 74.8%. [ABSTRACT FROM AUTHOR]

Summary: Probability and chance are essential concepts, not just in statistics but in real life. We present an adaptable activity which investigates what we mean by bias, how we can identify bias, and how we can use it to our advantage! [ABSTRACT FROM AUTHOR]

Objectives: The aim of this study is to analyze the data obtained from a randomized trial on the prevention of influenza by gargling with green tea, which gave nonsignificant results based on frequentist approaches, by using Bayesian approaches. Methods: The posterior proportion, with 95% credible interval (CrI), of influenza in each group was calculated. The Bayesian index θ is the probability that a hypothesis is true. In this case, θ is the probability that the hypothesis that green tea gargling reduced influenza compared with water gargling is true. Univariate and multivariate logistic regression analyses were also performed by using the Markov chain Monte Carlo method. Results: The full analysis set included 747 participants. During the study period, influenza occurred in 44 participants (5.9%). The difference between the two independent binominal proportions was −0.019 (95% CrI, −0.054 to 0.015; θ = 0.87). The partial regression coefficients in the univariate analysis were −0.35 (95% CrI, −1.00 to 0.24) with use of a uniform prior and −0.34 (95% CrI, −0.96 to 0.27) with use of a Jeffreys prior. In the multivariate analysis, the values were −0.37 (95% CrI, −0.96 to 0.30) and −0.36 (95% CrI, −1.03 to 0.21), respectively. Conclusions: The difference between the two independent binominal proportions was less than 0, and θ was greater than 0.85. Therefore, green tea gargling may slightly reduce influenza compared with water gargling. This analysis suggests that green tea gargling can be an additional preventive measure for use with other pharmaceutical and nonpharmaceutical measures and indicates the need for additional studies to confirm the effect of green tea gargling. [ABSTRACT FROM AUTHOR]

Single-value failure temperatures for fire loss of electrical cable functionality have been the norm for Fire Probabilistic Risk Assessments since the publication in 2005 of NUREG/CR-6850. If the calculated exposure temperature matches or exceeds the cable failure temperature, electrical failure is always assumed; if not, no failure is assumed. While this can be relaxed somewhat if a distribution for the exposure temperature is estimated, use of a distribution on the cable failure temperature itself more readily enables such relaxation and, therefore, a more realistic assessment. This paper develops probability distributions for different generic cable types (based on insulation) using data from the US Nuclear Regulatory Commission tests. Results indicate mean failure temperatures considerably higher than those used deterministically, 252°C, 421°C and 383°C, respectively for thermoplastic, thermoset and Kerite-FR. This suggests considerable relaxation from the conservatism inherent using the deterministic failure temperatures could be achieved. The paper then postulates two hypothetical distributions on the exposure temperature from applying a fire phenomenological model in a statistical way to estimate the possible relaxation using the distributed cable failure temperatures to enhance the realism of the assessment. Examples show that use of probabilistically-distributed cable failure temperatures (in conjunction with similar for exposure temperatures) can reduce the probability of electrical failure for a normally-distributed exposure temperature with a mean of 350°C and standard deviation of 58.3°C by factors of approximately three and eight for Kerite-FR and thermoset cables, respectively. The reduction would be less pronounced for thermoplastic cables, although larger reductions would be possible here as well for lower exposure temperatures (e.g., a factor of two). [ABSTRACT FROM AUTHOR]

Accurate measurement and a cutoff probability with inferential statistics are not wholly compatible. Fisher understood this when he developed the F test to deal with measurement variability and to make judgments on manipulations that may be worth further study. Neyman and Pearson focused on modeled distributions whose parameters were highly determined and concluded that inferential judgments following an F test could be made with accuracy because the distribution parameters were determined. Neyman and Pearson's approach in the application of statistical analyses using alpha and beta error rates has played a dominant role guiding inferential judgments, appropriately in highly determined situations and inappropriately in scientific exploration. Fisher tried to explain the different situations, but, in part due to some obscure wording, generated a long standing dispute that currently has left the importance of Fisher's p<.05 criteria not fully understood and a general endorsement of the Neyman and Pearson error rate approach. Problems were compounded with power calculations based on effect sizes following significant results entering into exploratory science. To understand in a practical sense when each approach should be used, a dimension reflecting varying levels of certainty or knowledge of population distributions is presented. The dimension provides a taxonomy of statistical situations and appropriate approaches by delineating four zones that represent how well the underlying population of interest is defined ranging from exploratory situations to highly determined populations. [ABSTRACT FROM AUTHOR]

The accumulated published database of stable lead isotope analyses of ore and slag specimens taken from Anatolian mining sites that parallel the Black Sea coast has been augmented with 22 additional analyses of such specimens carried out at the National Institute of Standards and Technology. Multivariate statistical analysis has been used to divide this composite database into five separate ore source groups. Evidence that most of these ore sources were exploited for the production of metal artefacts during the Bronze Age and Phrygian Period has been obtained by statistically comparing to them the isotope ratios of 184 analysed artefacts from nine archaeological sites situated within a few hundred kilometres of these mining sites. Also, Appendix B contains 36 new isotope analyses of ore specimens from Central Taurus mining sites that are compatible with and augment the four Central Taurus Ore Source Groups defined in Yener et al. (1991). [ABSTRACT FROM AUTHOR]