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.

Notes on PCA, Regularization, Sparsity and Support Vector Machines

We derive a new representation for a function as a linear combination of local correlation kernels at optimal sparse locations and discuss its relation to PCA, regularization, sparsity principles and Support Vector Machines. We first review previous results for the approximation of a function from discrete data (Girosi, 1998) in the context of Vapnik"s feature space and dual representation (Vapnik, 1995). We apply them to show 1) that a standard regularization functional with a stabilizer defined in terms of the correlation function induces a regression function in the span of the feature space of classical Principal Components and 2) that there exist a dual representations of the regression function in terms of a regularization network with a kernel equal to a generalized correlation function. We then describe the main observation of the paper: the dual representation in terms of the correlation function can be sparsified using the Support Vector Machines (Vapnik, 1982) technique and this operation is equivalent to sparsify a large dictionary of basis functions adapted to the task, using a variation of Basis Pursuit De-Noising (Chen, Donoho and Saunders, 1995; see also related work by Donahue and Geiger, 1994; Olshausen and Field, 1995; Lewicki and Sejnowski, 1998). In addition to extending the close relations between regularization, Support Vector Machines and sparsity, our work also illuminates and formalizes the LFA concept of Penev and Atick (1996). We discuss the relation between our results, which are about regression, and the different problem of pattern classification.

On the Noise Model of Support Vector Machine Regression

Support Vector Machines Regression (SVMR) is a regression technique which has been recently introduced by V. Vapnik and his collaborators (Vapnik, 1995; Vapnik, Golowich and Smola, 1996). In SVMR the goodness of fit is measured not by the usual quadratic loss function (the mean square error), but by a different loss function called Vapnik"s $epsilon$- insensitive loss function, which is similar to the "robust" loss functions introduced by Huber (Huber, 1981). The quadratic loss function is well justified under the assumption of Gaussian additive noise. However, the noise model underlying the choice of Vapnik's loss function is less clear. In this paper the use of Vapnik's loss function is shown to be equivalent to a model of additive and Gaussian noise, where the variance and mean of the Gaussian are random variables. The probability distributions for the variance and mean will be stated explicitly. While this work is presented in the framework of SVMR, it can be extended to justify non-quadratic loss functions in any Maximum Likelihood or Maximum A Posteriori approach. It applies not only to Vapnik's loss function, but to a much broader class of loss functions.

Comparing methods for dimensionality reduction when data are density functions

Functional Data Analysis (FDA) deals with samples where a whole function is observed for each individual. A particular case of FDA is when the observed functions are density functions, that are also an example of infinite dimensional compositional data. In this work we compare several methods for dimensionality reduction for this particular type of data: functional principal components analysis (PCA) with or without a previous data transformation and multidimensional scaling (MDS) for diferent inter-densities distances, one of them taking into account the compositional nature of density functions. The difeerent methods are applied to both artificial and real data (households income distributions)

Executive functions in school-aged children: age and socioeconomic status effects

Lately, the study of prefrontal executive functions in grade scholars has noticeably increased. The aim of this study is to investigate the influence of age and socioeconomic status (sEs) on executive tasks performance and to analyze those socioeconomic variables that predict a better execution. A sample of 254 children aged between 7 and 12 years from the city of santa Fe, Argentina and belonging to different socioeconomic status were tested. A bat- tery of executive functions sensitive to prefrontal function was used to obtain the results. These in- dicate a significant influence of age and SES on executive functions. The cognitive patterns follow a different path according to the development and sEs effect. Besides, it is revealed a pattern of low cognitive functioning in low-sEs children in all executive functions. Finally, from the variables included in this study, it was found that only the educational level of the mother and the housing conditions are associated to the children’s executive function. The results are discussed in terms of the influence of the cerebral maturation and the envi- ronmental variables in the executive functioning.

Distribución hiperbólica generalizada: una aplicación en la selección de portafolios y en la cuantificación de medidas de riesgo de mercado

En este trabajo se implementa una metodología para incluir momentos de orden superior en la selección de portafolios, haciendo uso de la Distribución Hiperbólica Generalizada, para posteriormente hacer un análisis comparativo frente al modelo de Markowitz.

Espacios locales de participación y construcción de ciudadanía un estudio de caso de la Junta Administradora Local de Teusaquillo

La propuesta de la investigación consiste en observar y comprobar si efectivamente los espacios de participación local surgidos después de la constitución de 1991, en un esfuerzo del estado colombiano y la administración Distrital por descentralizar la función administrativa, han sido efectivos dentro del proceso de construcción de ciudadanía, y en qué forma los ediles interactúan con los habitantes de la localidad y canalizan las demandas de la comunidad. Por lo mismo, se pretende indagar sobre las características de la participación ciudadana en la localidad y la aparición de tensiones entre la apertura de espacios de participación, y el uso de ellos por redes políticas locales de tipo clientelista. De esta forma, se busca comprobar si surge así, un fenómeno de construcción de ciudadanía permeado por intereses de carácter particularista, limitando a los ciudadanos, que como miembros de las redes de intermediación política, acceden a los derechos de ciudadanía que ofrecen los nuevos espacios de representación

Generalized Higher-Order Voronoi Diagrams on Polyhedral Surfaces

We present an algorithm for computing exact shortest paths, and consequently distances, from a generalized source (point, segment, polygonal chain or polygonal region) on a possibly non-convex polyhedral surface in which polygonal chain or polygon obstacles are allowed. We also present algorithms for computing discrete Voronoi diagrams of a set of generalized sites (points, segments, polygonal chains or polygons) on a polyhedral surface with obstacles. To obtain the discrete Voronoi diagrams our algorithms, exploiting hardware graphics capabilities, compute shortest path distances defined by the sites

The mapping of the local contributions of Fermi and Coulomb correlation into intracule and extracule density distributions

The contributions of the correlated and uncorrelated components of the electron-pair density to atomic and molecular intracule I(r) and extracule E(R) densities and its Laplacian functions ∇2I(r) and ∇2E(R) are analyzed at the Hartree-Fock (HF) and configuration interaction (CI) levels of theory. The topologies of the uncorrelated components of these functions can be rationalized in terms of the corresponding one-electron densities. In contrast, by analyzing the correlated components of I(r) and E(R), namely, IC(r) and EC(R), the effect of electron Fermi and Coulomb correlation can be assessed at the HF and CI levels of theory. Moreover, the contribution of Coulomb correlation can be isolated by means of difference maps between IC(r) and EC(R) distributions calculated at the two levels of theory. As application examples, the He, Ne, and Ar atomic series, the C2-2, N2, O2+2 molecular series, and the C2H4 molecule have been investigated. For these atoms and molecules, it is found that Fermi correlation accounts for the main characteristics of IC(r) and EC(R), with Coulomb correlation increasing slightly the locality of these functions at the CI level of theory. Furthermore, IC(r), EC(R), and the associated Laplacian functions, reveal the short-ranged nature and high isotropy of Fermi and Coulomb correlation in atoms and molecules

Mosteiro de S. Salvador de Paço de Sousa Contributo para a sua reabilitação arquitetónica

Este trabalho pretende contribuir para a reabilitação do Mosteiro de S. Salvador de Paço de Sousa assente na interpretação espacial do complexo monástico desde a fundação à extinção da comunidade religiosa, com particular nota para a intervenção efetuada pela DGEMN - Direção Geral dos Edifícios e Monumentos Nacionais, no segundo quartel do século XX. Procura trazer pela primeira vez à luz uma representação dos seus espaços, usos e funções ao longo do tempo com recurso a desenhos, esquemas e imagens para permitir a apreensão do conjunto em determinadas fases da história e por consequência fundamentar uma eventual intervenção. O estudo fundamenta-se na recolhe de dados escritos que permitam aferir a vivência dos espaços, sua simbologia e utilidade, nomeadamente através de referências históricas e de carater simbólico e documentação gráfica que permita avaliar e comparar o assunto de estudo com outros exemplos e situações análogas. O cruzamento de dados analisados resulta na sua simbolização gráfica, de forma a veicular o entendimento espacial e simbólico, que remete o real conhecimento da arquitetura do Mosteiro de S. Salvador de Paço de Sousa e, por comparação, viabiliza um ponto de partida para futuras intervenções, seja deste mosteiro, seja de edifícios de características semelhantes.

Realizability of Point Processes

There are various situations in which it is natural to ask whether a given collection of k functions, ρ j (r 1,…,r j ), j=1,…,k, defined on a set X, are the first k correlation functions of a point process on X. Here we describe some necessary and sufficient conditions on the ρ j ’s for this to be true. Our primary examples are X=ℝ d , X=ℤ d , and X an arbitrary finite set. In particular, we extend a result by Ambartzumian and Sukiasian showing realizability at sufficiently small densities ρ 1(r). Typically if any realizing process exists there will be many (even an uncountable number); in this case we prove, when X is a finite set, the existence of a realizing Gibbs measure with k body potentials which maximizes the entropy among all realizing measures. We also investigate in detail a simple example in which a uniform density ρ and translation invariant ρ 2 are specified on ℤ; there is a gap between our best upper bound on possible values of ρ and the largest ρ for which realizability can be established.

Linear viscoelasticity from molecular dynamics simulation of entangled polymers

The linear viscoelastic (LVE) spectrum is one of the primary fingerprints of polymer solutions and melts, carrying information about most relaxation processes in the system. Many single chain theories and models start with predicting the LVE spectrum to validate their assumptions. However, until now, no reliable linear stress relaxation data were available from simulations of multichain systems. In this work, we propose a new efficient way to calculate a wide variety of correlation functions and mean-square displacements during simulations without significant additional CPU cost. Using this method, we calculate stress−stress autocorrelation functions for a simple bead−spring model of polymer melt for a wide range of chain lengths, densities, temperatures, and chain stiffnesses. The obtained stress−stress autocorrelation functions were compared with the single chain slip−spring model in order to obtain entanglement related parameters, such as the plateau modulus or the molecular weight between entanglements. Then, the dependence of the plateau modulus on the packing length is discussed. We have also identified three different contributions to the stress relaxation:  bond length relaxation, colloidal and polymeric. Their dependence on the density and the temperature is demonstrated for short unentangled systems without inertia.

Driving places - Marc Augé, non-places, and the geographies of England's M1 motorway

In this article I provide a critical account of the 'placing' of England's M1 motor-way. I start by critiquing Marc Auge's anthropological writings on 'non-places' which have provided a common point of reference for academics discussing spaces of travel, consumption and exchange in the contemporary world. I argue that Auge's ethnology of supermodernity results in a rather partial account of these sites, that he overstates the novelty of contemporary experiences of these spaces, and that he fails to acknowledge the heterogeneity and materiality of the social networks bound up with the production of non-places/places. I suggest that, rather than focusing on the presences and absences associated with the polarities of place and non-place, academics should examine the multiple, partial, dynamic and relational 'placings' which arise through the diverse performances and movements associated with travel, consumption and exchange. I then trace the topologies of England's M1 motorway, examining some of the different ways in which the motorway has been assembled, performed and placed over the past 45 years.

'Operation motorway': landscapes of construction on England's M1 motorway

Recent years have seen an increasing number of academics attempt to write more process-oriented and 'nonrepresentational' accounts of landscape. Drawing upon this literature, I discuss a number of the movements, materialities, and practices entailed in constructing England's M1 motorway in the late 1950s. The performances, movements and durability of a diverse range of things-including earth-moving machines, public relations brochures, maps, helicopters, senior engineers, aggregate and labourers-are shown to be important to the construction and ordering of the motorway and spaces of the construction company in different times and spaces, with people's experiences or understandings of construction, both now and in the past, emerging through memories, talk and embodied encounters with architectures, texts and artefacts which are assembled, circulated and/or archived. Aerial perspectives assumed a prominent role in depictions of construction, while journalists and engineers frequently drew upon a military vocabulary and alluded to the military nature of the project when discussing the motorway. (c) 2005 Elsevier Ltd. All rights reserved.

A review of studies performed to assess metal uptake by earthworms

Earthworms perform a number of essential functions in soil; the impacts of metals on earthworms are often investigated. In this review we consider the range of earthworm species, types of soil and forms of metal for which metal uptake and accumulation have been studied, the design of these experiments and the quantitative relationships that have been derived to predict earthworm metal body burden. We conclude that there is a need for more studies on earthworm species other than Eisenia fetida in order to apply the large existing database on this earthworm to other, soil dwelling species. To aid comparisons between studies agreement is needed on standard protocols that define exposure and deputation periods and the parameters, such as soil solution composition, soil chemical and physical properties to be measured. It is recommended that more field or terrestrial model ecosystem studies using real contaminated soil rather than metal-amended artificial soils are performed. (c) 2006 Elsevier Ltd. All rights reserved.