Automatic Detection of Non-deverbal Event Nouns for Quick Lexicon Production

In this work we present the results of experimental work on the development of lexical class-based lexica by automatic means. Our purpose is to assess the use of linguistic lexical-class based information as a feature selection methodology for the use of classifiers in quick lexical development. The results show that the approach can help reduce the human effort required in the development of language resources significantly.

Dynamics of the two-dimensional gonihedric spin model

In this paper, we study dynamical aspects of the two-dimensional (2D) gonihedric spin model using both numerical and analytical methods. This spin model has vanishing microscopic surface tension and it actually describes an ensemble of loops living on a 2D surface. The self-avoidance of loops is parametrized by a parameter ¿. The ¿=0 model can be mapped to one of the six-vertex models discussed by Baxter, and it does not have critical behavior. We have found that allowing for ¿¿0 does not lead to critical behavior either. Finite-size effects are rather severe, and in order to understand these effects, a finite-volume calculation for non-self-avoiding loops is presented. This model, like his 3D counterpart, exhibits very slow dynamics, but a careful analysis of dynamical observables reveals nonglassy evolution (unlike its 3D counterpart). We find, also in this ¿=0 case, the law that governs the long-time, low-temperature evolution of the system, through a dual description in terms of defects. A power, rather than logarithmic, law for the approach to equilibrium has been found.

Coarsening of solid-liquid mixtures in a random acceleration field

The effects of flow induced by a random acceleration field (g-jitter) are considered in two related situations that are of interest for microgravity fluid experiments: the random motion of isolated buoyant particles, and diffusion driven coarsening of a solid-liquid mixture. We start by analyzing in detail actual accelerometer data gathered during a recent microgravity mission, and obtain the values of the parameters defining a previously introduced stochastic model of this acceleration field. The diffusive motion of a single solid particle suspended in an incompressible fluid that is subjected to such random accelerations is considered, and mean squared velocities and effective diffusion coefficients are explicitly given. We next study the flow induced by an ensemble of such particles, and show the existence of a hydrodynamically induced attraction between pairs of particles at distances large compared with their radii, and repulsion at short distances. Finally, a mean field analysis is used to estimate the effect of g-jitter on diffusion controlled coarsening of a solid-liquid mixture. Corrections to classical coarsening rates due to the induced fluid motion are calculated, and estimates are given for coarsening of Sn-rich particles in a Sn-Pb eutectic fluid, an experiment to be conducted in microgravity in the near future.

Multifractal wavelet filter of natural images

Natural images are characterized by the multiscaling properties of their contrast gradient, in addition to their power spectrum. In this Letter we show that those properties uniquely define an intrinsic wavelet and present a suitable technique to obtain it from an ensemble of images. Once this wavelet is known, images can be represented as expansions in the associated wavelet basis. The resulting code has the remarkable properties that it separates independent features at different resolution level, reducing the redundancy, and remains essentially unchanged under changes in the power spectrum. The possible generalization of this representation to other systems is discussed.

A model of the wind direction signature in the Stokes emission vector from the ocean surface at microwave frequencies

This paper presents a model of the Stokes emission vector from the ocean surface. The ocean surface is described as an ensemble of facets with Cox and Munk's (1954) Gram-Charlier slope distribution. The study discusses the impact of different up-wind and cross-wind rms slopes, skewness, peakedness, foam cover models and atmospheric effects on the azimuthal variation of the Stokes vector, as well as the limitations of the model. Simulation results compare favorably, both in mean value and azimuthal dependence, with SSM/I data at 53° incidence angle and with JPL's WINDRAD measurements at incidence angles from 30° to 65°, and at wind speeds from 2.5 to 11 m/s.

Average performance analysis of circular and hyperbolic geolocation

A comparative performance analysis of four geolocation methods in terms of their theoretical root mean square positioning errors is provided. Comparison is established in two different ways: strict and average. In the strict type, methods are examined for a particular geometric configuration of base stations(BSs) with respect to mobile position, which determines a givennoise profile affecting the respective time-of-arrival (TOA) or timedifference-of-arrival (TDOA) estimates. In the average type, methodsare evaluated in terms of the expected covariance matrix ofthe position error over an ensemble of random geometries, so thatcomparison is geometry independent. Exact semianalytical equationsand associated lower bounds (depending solely on the noiseprofile) are obtained for the average covariance matrix of the positionerror in terms of the so-called information matrix specific toeach geolocation method. Statistical channel models inferred fromfield trials are used to define realistic prior probabilities for therandom geometries. A final evaluation provides extensive resultsrelating the expected position error to channel model parametersand the number of base stations.

Estimating 750 years of temperature variations and uncertainties in the Pyrenees by tree-ring reconstructions and climate simulations.

Past temperature variations are usually inferred from proxy data or estimated using general circulation models. Comparisons between climate estimations derived from proxy records and from model simulations help to better understand mechanisms driving climate variations, and also offer the possibility to identify deficiencies in both approaches. This paper presents regional temperature reconstructions based on tree-ring maximum density series in the Pyrenees, and compares them with the output of global simulations for this region and with regional climate model simulations conducted for the target region. An ensemble of 24 reconstructions of May-to-September regional mean temperature was derived from 22 maximum density tree-ring site chronologies distributed over the larger Pyrenees area. Four different tree-ring series standardization procedures were applied, combining two detrending methods: 300-yr spline and the regional curve standardization (RCS). Additionally, different methodological variants for the regional chronology were generated by using three different aggregation methods. Calibration verification trials were performed in split periods and using two methods: regression and a simple variance matching. The resulting set of temperature reconstructions was compared with climate simulations performed with global (ECHO-G) and regional (MM5) climate models. The 24 variants of May-to-September temperature reconstructions reveal a generally coherent pattern of inter-annual to multi-centennial temperature variations in the Pyrenees region for the last 750 yr. However, some reconstructions display a marked positive trend for the entire length of the reconstruction, pointing out that the application of the RCS method to a suboptimal set of samples may lead to unreliable results. Climate model simulations agree with the tree-ring based reconstructions at multi-decadal time scales, suggesting solar variability and volcanism as the main factors controlling preindustrial mean temperature variations in the Pyrenees. Nevertheless, the comparison also highlights differences with the reconstructions, mainly in the amplitude of past temperature variations and in the 20th century trends. Neither proxy-based reconstructions nor model simulations are able to perfectly track the temperature variations of the instrumental record, suggesting that both approximations still need further improvements.

Classifier combination for in vivo magnetic resonance spectra of brain tumours

In this paper we present a multi-stage classifier for magnetic resonance spectra of human brain tumours which is being developed as part of a decision support system for radiologists. The basic idea is to decompose a complex classification scheme into a sequence of classifiers, each specialising in different classes of tumours and trying to reproducepart of the WHO classification hierarchy. Each stage uses a particular set of classification features, which are selected using a combination of classical statistical analysis, splitting performance and previous knowledge.Classifiers with different behaviour are combined using a simple voting scheme in order to extract different error patterns: LDA, decision trees and the k-NN classifier. A special label named "unknown¿ is used when the outcomes of the different classifiers disagree. Cascading is alsoused to incorporate class distances computed using LDA into decision trees. Both cascading and voting are effective tools to improve classification accuracy. Experiments also show that it is possible to extract useful information from the classification process itself in order to helpusers (clinicians and radiologists) to make more accurate predictions and reduce the number of possible classification mistakes.

Asymptotic flocking dynamics for the kinetic Cucker-Smale model

In this paper, we analyse the asymptotic behavior of solutions of the continuous kinetic version of flocking by Cucker and Smale [16], which describes the collective behavior of an ensemble of organisms, animals or devices. This kinetic version introduced in [24] is here obtained starting from a Boltzmann-type equation. The large-time behavior of the distribution in phase space is subsequently studied by means of particle approximations and a stability property in distances between measures. A continuous analogue of the theorems of [16] is shown to hold for the solutions on the kinetic model. More precisely, the solutions will concentrate exponentially fast their velocity to their mean while in space they will converge towards a translational flocking solution.

Hierarchical conditional random fields for parts based models matching

A parts based model is a parametrization of an object class using a collection of landmarks following the object structure. The matching of parts based models is one of the problems where pairwise Conditional Random Fields have been successfully applied. The main reason of their effectiveness is tractable inference and learning due to the simplicity of involved graphs, usually trees. However, these models do not consider possible patterns of statistics among sets of landmarks, and thus they sufffer from using too myopic information. To overcome this limitation, we propoese a novel structure based on a hierarchical Conditional Random Fields, which we explain in the first part of this memory. We build a hierarchy of combinations of landmarks, where matching is performed taking into account the whole hierarchy. To preserve tractable inference we effectively sample the label set. We test our method on facial feature selection and human pose estimation on two challenging datasets: Buffy and MultiPIE. In the second part of this memory, we present a novel approach to multiple kernel combination that relies on stacked classification. This method can be used to evaluate the landmarks of the parts-based model approach. Our method is based on combining responses of a set of independent classifiers for each individual kernel. Unlike earlier approaches that linearly combine kernel responses, our approach uses them as inputs to another set of classifiers. We will show that we outperform state-of-the-art methods on most of the standard benchmark datasets.

A Novel Breast Tissue Density Classification Methodology

It has been shown that the accuracy of mammographic abnormality detection methods is strongly dependent on the breast tissue characteristics, where a dense breast drastically reduces detection sensitivity. In addition, breast tissue density is widely accepted to be an important risk indicator for the development of breast cancer. Here, we describe the development of an automatic breast tissue classification methodology, which can be summarized in a number of distinct steps: 1) the segmentation of the breast area into fatty versus dense mammographic tissue; 2) the extraction of morphological and texture features from the segmented breast areas; and 3) the use of a Bayesian combination of a number of classifiers. The evaluation, based on a large number of cases from two different mammographic data sets, shows a strong correlation ( and 0.67 for the two data sets) between automatic and expert-based Breast Imaging Reporting and Data System mammographic density assessment

Many worlds in one

A generic prediction of inflation is that the thermalized region we inhabit is spatially infinite. Thus, it contains an infinite number of regions of the same size as our observable universe, which we shall denote as O regions. We argue that the number of possible histories which may take place inside of an O region, from the time of recombination up to the present time, is finite. Hence, there are an infinite number of O regions with identical histories up to the present, but which need not be identical in the future. Moreover, all histories which are not forbidden by conservation laws will occur in a finite fraction of all O regions. The ensemble of O regions is reminiscent of the ensemble of universes in the many-world picture of quantum mechanics. An important difference, however, is that other O regions are unquestionably real.

La lectura compartida: una eina per l'aprenentatge de la lecto-escriptura en equips cooperatius

L’estudi de cas que es presenta l’hem realitzat a Cicle inicial (CI) de l’escola els Castanyers de Viladrau. Hem portat a terme un projecte per tal d’estudiar la lectura compartida tot utilitzant equips cooperatius (Pujolàs, 2008). Els objectius d’aquest treball són: en primer lloc, ensenyar els alumnes a aprendre de manera cooperativa, és a dir, saber treballar conjuntament, tenir respecte mutu, escoltar als companys, proporcionar ajudes a la resta de l’equip i respectar els punts de vista dels altres, i, en segon lloc, millorar la comprensió dels texts; millorar la lectura i l’escriptura, saber realitzar les tasques i rols, saber seleccionar informació, aprendre a realitzar preguntes. Per portar-lo a terme vam realitzar set sessions en les qual la lectura compartida i els equips cooperatius eren les bases de totes elles. El conjunts de les activitats ens ha permès evidenciar una cohesió del grup-classe més elevada que abans de realitzar-les, així com una lleugera millora en els aspectes relacionats amb la comprensió lectora.

Redrawing Tarraco

The city of Tarragona houses an important architectural heritage mainly from its past as ‘Tarraco’, capital of the Roman province of Hispania Citerior, but also from its medieval and late 19th century history. The archaeological ensemble of Tarraco was inscribed as a UNESCO World Heritage Site in 2000, but although many efforts have been devoted by archaeologists and historians to unveil and understand the history and aspect of the Roman city, many aspects remain unknown. This is largely caused by the absence of a coherent body of historiographical material, which is todays cattered across several institutions and, specially, the lack of precise and useful graphical representations of the remains and of the existing city that allows in-depth analysis and interpretations of future findings. In recent years, researchers from the Catalan Institute of Classical Archaeology (ICAC) and the Architecture School of the URV (ETSA) have teamed up to produce comprehensive, detailed graphic materials, including a new set of plans and sections of the old city, of the grandiose areas of representation of the Provincial capital, and of the hidden structures beneath the city’s surface. These have been executed with the latest technologies (fotogrammetry, laser scanning) but also with traditional methods (measurement, topography), on t op of a mixture of existing materials (hand-drafted cartography from municipal master plans) and of historical and archaeological documentation.

Textural classification of land cover using support vector machines: an empirical comparison with parametric, non parametric and hybrid classifiers in the Bolivian Amazon

Land cover classification is a key research field in remote sensing and land change science as thematic maps derived from remotely sensed data have become the basis for analyzing many socio-ecological issues. However, land cover classification remains a difficult task and it is especially challenging in heterogeneous tropical landscapes where nonetheless such maps are of great importance. The present study aims to establish an efficient classification approach to accurately map all broad land cover classes in a large, heterogeneous tropical area of Bolivia, as a basis for further studies (e.g., land cover-land use change). Specifically, we compare the performance of parametric (maximum likelihood), non-parametric (k-nearest neighbour and four different support vector machines - SVM), and hybrid classifiers, using both hard and soft (fuzzy) accuracy assessments. In addition, we test whether the inclusion of a textural index (homogeneity) in the classifications improves their performance. We classified Landsat imagery for two dates corresponding to dry and wet seasons and found that non-parametric, and particularly SVM classifiers, outperformed both parametric and hybrid classifiers. We also found that the use of the homogeneity index along with reflectance bands significantly increased the overall accuracy of all the classifications, but particularly of SVM algorithms. We observed that improvements in producer’s and user’s accuracies through the inclusion of the homogeneity index were different depending on land cover classes. Earlygrowth/degraded forests, pastures, grasslands and savanna were the classes most improved, especially with the SVM radial basis function and SVM sigmoid classifiers, though with both classifiers all land cover classes were mapped with producer’s and user’s accuracies of around 90%. Our approach seems very well suited to accurately map land cover in tropical regions, thus having the potential to contribute to conservation initiatives, climate change mitigation schemes such as REDD+, and rural development policies.