A Note on the Generalization Performance of Kernel Classifiers with Margin

We present distribution independent bounds on the generalization misclassification performance of a family of kernel classifiers with margin. Support Vector Machine classifiers (SVM) stem out of this class of machines. The bounds are derived through computations of the $V_gamma$ dimension of a family of loss functions where the SVM one belongs to. Bounds that use functions of margin distributions (i.e. functions of the slack variables of SVM) are derived.

Sparse Correlation Kernel Analysis and Reconstruction

This paper presents a new paradigm for signal reconstruction and superresolution, Correlation Kernel Analysis (CKA), that is based on the selection of a sparse set of bases from a large dictionary of class- specific basis functions. The basis functions that we use are the correlation functions of the class of signals we are analyzing. To choose the appropriate features from this large dictionary, we use Support Vector Machine (SVM) regression and compare this to traditional Principal Component Analysis (PCA) for the tasks of signal reconstruction, superresolution, and compression. The testbed we use in this paper is a set of images of pedestrians. This paper also presents results of experiments in which we use a dictionary of multiscale basis functions and then use Basis Pursuit De-Noising to obtain a sparse, multiscale approximation of a signal. The results are analyzed and we conclude that 1) when used with a sparse representation technique, the correlation function is an effective kernel for image reconstruction and superresolution, 2) for image compression, PCA and SVM have different tradeoffs, depending on the particular metric that is used to evaluate the results, 3) in sparse representation techniques, L_1 is not a good proxy for the true measure of sparsity, L_0, and 4) the L_epsilon norm may be a better error metric for image reconstruction and compression than the L_2 norm, though the exact psychophysical metric should take into account high order structure in images.

On the V(subscript gamma) Dimension for Regression in Reproducing Kernel Hilbert Spaces

This paper presents a computation of the $V_gamma$ dimension for regression in bounded subspaces of Reproducing Kernel Hilbert Spaces (RKHS) for the Support Vector Machine (SVM) regression $epsilon$-insensitive loss function, and general $L_p$ loss functions. Finiteness of the RV_gamma$ dimension is shown, which also proves uniform convergence in probability for regression machines in RKHS subspaces that use the $L_epsilon$ or general $L_p$ loss functions. This paper presenta a novel proof of this result also for the case that a bias is added to the functions in the RKHS.

A comparison of the alr and ilr transformations for kernel density estimation of compositional data

In a seminal paper, Aitchison and Lauder (1985) introduced classical kernel density estimation techniques in the context of compositional data analysis. Indeed, they gave two options for the choice of the kernel to be used in the kernel estimator. One of these kernels is based on the use the alr transformation on the simplex SD jointly with the normal distribution on RD-1. However, these authors themselves recognized that this method has some deficiencies. A method for overcoming these dificulties based on recent developments for compositional data analysis and multivariate kernel estimation theory, combining the ilr transformation with the use of the normal density with a full bandwidth matrix, was recently proposed in Martín-Fernández, Chacón and Mateu- Figueras (2006). Here we present an extensive simulation study that compares both methods in practice, thus exploring the finite-sample behaviour of both estimators

Medición de la carga de enfermedad en una Entidad Promotora de Salud de Colombia año 2008

Introducción: los AVAD (años de vida ajustados a discapacidad) son la unidad de medida de la carga de enfermedad, permiten estimar las pérdidas de salud para una población con respecto a las consecuencias mortales y no mortales de las enfermedades. Metodología: se siguió la metodología del estudio de Carga Global de Enfermedad 2001. Para cálcular los AVAD´s se generan primero sus insumos: Años Vividos con Discapacidad y Años de Vida perdidos por muerte Prematura para lo cual se realizó una tabla de vida actual para la mortalidad general y se tomó la prevalencia de las enfermedades para la morbilidad. Resultados: la tasa de AVAD´s estimada en este estudio fue de 246 AVAD´s por mil personas, correspondientes a 366678 AVAD´s totales, concentrados en su mayoría en las enfermedades del grupo II con un total de 228 AVAD´s por mil habitantes, seguido del grupo I con 17,3 AVAD´s por mil habitantes; y con la menor participación, el grupo III con 0,6 AVAD´s. Los AVAD´s dependientes de la mortalidad fueron 2,4 por mil personas y 244 por discapacidad. La Enfermedad Pulmonar Obstructiva Crónica en ambos géneros ocupa el primer lugar como enfermedad generadora de AVAD´s. Discusión: el 17% de AVAD´s en este estudio pertenecen al grupo I, cifra que se aproxima a la obtenida en nuestro país. El menor peso porcentual se encuentra en el número de AVAD´s del grupo III tanto para Colombia como para la E.P.S. Las limitaciones del estudio se relacionan con la calidad de los datos de los registros de mortalidad.

Psychometric properties of the scale NBAS applied to preterm or low birth weight

Objective: to assess, in a randomized sample of measurements made with the NBAS scale, in preterm and / or low birth weight babies, some psychometric characteristics of the same as her construct validity and internal consistency. Materials and methods: we designed a study of assessment of Psychometric properties. From a study of cross-sectional measurements, there were made assessment of the statistic behavior of the items, individually and in their internal consistency, then, there were carried out, type exploratory and confirmatory factor analysis. Results: we find that, in the analysis of preterm or low birth weight babies, the scale remains largely, the overall structure of factors proposed since 1982, in relation to the dimensions of habituation, social interactive and complementary. The items that make up the dimensions of SNA and motor systems are integrated. Also, we found Cronbach alpha values for internal consistency, which reflects higher to medium correlation levels in the behavioral and reflex items. Conclusion: the NBAS scale reflects characteristics of construct validity and internal consistency, which credits it as a useful tool in clinical assessment of neurodevelopmental effects of preterm or low birth weight babies. It is necessary to analyze the structure of the scale when used for purposes of scientific research.

Morbilidad y mortalidad en prematuros menores de 1500 gramos en un hospital regional del 2011 a 2013

Introducción: El presente estudio pretende determinar la mortalidad y caracterizar morbilidad de este grupo de recién nacidos, para establecer planes de mejoramiento. Materiales y método: Estudio descriptivo retrospectivo de corte transversal. Se revisaron 158 historias clínicas de los recién nacidos prematuros menores de 1500 gramos hospitalizados en la unidad de cuidados intensivos neonatales del Hospital Universitario Departamental de Nariño durante el periodo 2011 al 2013. La información fue analizada estadísticamente. Resultados: Se encontró que de 5447 nacidos vivos el 2,9 % fueron menores de 1500 gramos. 52,5 % eran de género masculino, 63,9% nacieron por cesárea. El 23,4 % no recibió esteroides antenatales. La tasa de mortalidad para el periodo de estudio en este grupo de pacientes fue de 7.3 por mil nacidos vivos. El 100% de los recién nacidos de menos de 750 gr fallecieron. Mientras que no se registro ninguna sobrevida de menos de 24 semanas. Conclusiones: Podría establecerse este como límite de viabilidad el peso al nacer < 750 gr y menos de 24 semanas de gestación, en donde la muerte es prácticamente la regla, siempre teniendo en cuenta evaluar cuidadosamente cada caso particular. La morbilidad de los prematuros de muy bajo peso al nacer esta en los rangos reportados en la literatura.

The language used in early mother-child interaction by very-low-birth-weight children, with a focus on the emergence of grammar. 'El lenguaje utilizado por niños con muy bajo peso al nacer en la interacción temprana entre madre e hijo, con atención especial a la aparición de la gramática'.

Resumen basado en el de la publicación. Resumen en español

ABR as a predictor of neurodevelopment in very low birth weight neonates

This paper discusses a study to determine whether the neonatal ABR predicts neurodevelopmental delays in low birth weight infants.

General relationships between pressure, weight, and mass of a hydrostatic fluid

In curved geometries the hydrostatic pressure in a fluid does not equal the weight per unit area of the fluid above it. General weight–pressure and mass–pressure relationships for hydrostatic fluids in any geometry are derived. As an example of the mass–pressure relationship, we find a geometric reduction in surface pressure as large as 5 mbar on Earth and 39 mbar on Titan. We also present a thermodynamic interpretation of the geometric correction which, as a corollary, provides an independent proof of the hydrostatic relationship for general geometries.

The imminent availability of a potent weight-loss medicine in pharmacies: public perception of effectiveness

Orlistat is an effective weight-loss medicine, which will soon be available for purchase in pharmacies. We used a factorial experiment and found that informing people about the availability for purchase of this medicinal product previously only available on prescription resulted in higher ratings of perceived value and effectiveness compared to a natural health supplement even though we used the same statement about effectiveness. This positive perception of orlistat was not impaired by the provision of side-effect information. Orlistat will soon be available in pharmacies. Health professionals must act to prevent its misuse by those not overweight.

Characterisation of an s-type low molecular weight glutenin subunit of wheat and its proline and glutamine-rich repetitive domain

The low molecular weight glutenin subunits (LMW-GS) are major components of the glutenin polymers which determine the elastomeric properties of wheat (Triticum aestivum L.) gluten and dough. They comprise a complex mixture of components and have proved to be difficult to purify for detailed characterisation. The mature LMW subunit proteins comprise two structural domains, with one domain consisting of repeated sequences based on short peptide motifs. DNA sequences encoding this domain and a whole subunit were expressed in Escherichia coli and the recombinant proteins purified. Detailed comparisons by spectroscopy (CD, FT-IR) and dynamic light scattering indicated that the repetitive and non-repetitive domains of the proteins formed different structures with the former having an extended conformation with an equilibrium between poly-L-proline II-like structure and type II’ b-turns, and the latter a more compact globular structure rich in a-helix. Although the structures of these two domains appear to form independently, dynamic light scattering of the whole subunit dissolved in trifluoroethanol(TFE) suggested that they interact, leading to a more compact conformation. These observations may have relevance to the role of the LMW-GS in gluten structure and functionality.

Intake, live-weight gain and carcass characteristics of beef cattle given diets based on forage maize silage harvested at different stages of maturity

Advancing maturity of forage maize is associated with increases in the proportion of dry matter (DM) and starch, and decreases in the proportions of structural carbohydrates in the ensiled crop. This experiment investigated the effects of three maize silages of 291 (low), 339 (medium) and 393 (high) g DM per kg fresh weight on the performance of 48 Simmental. Holstein-Friesian cattle. Equal numbers of steers (mean start weight = 503 (s.d. 31.3) kg) and heifers (mean start weight = 378 (s.d. 11.2) kg) were offered individually isonitrogenous diets composed of the three silages plus a protein supplement with minerals once daily until slaughter at the target live weight of 575 and 475 kg for steers and heifers, respectively. Intake was reduced on the low diet (P < 0.01) compared with the other two treatments. Dietary starch intake increased by a total of 1 kg/day between low and medium diets but by only 0.2 kg/day between medium and high diets. Unlike starch intake, total neutral-detergent fibre intake showed no significant difference (P > 0.05) between diets. There were no differences in live-weight gain between treatments but differences (P < 0.05) in food conversion efficiency indicated relative gains of 115, 100 and 102 g gain per kg DM intake for diets low, medium and high, respectively. There were no differences between diets in carcass weights, fat score and overall conformation.