Data-driven clustering : new methods and applications

Abstract : This work is concerned with the development and application of novel unsupervised learning methods, having in mind two target applications: the analysis of forensic case data and the classification of remote sensing images. First, a method based on a symbolic optimization of the inter-sample distance measure is proposed to improve the flexibility of spectral clustering algorithms, and applied to the problem of forensic case data. This distance is optimized using a loss function related to the preservation of neighborhood structure between the input space and the space of principal components, and solutions are found using genetic programming. Results are compared to a variety of state-of--the-art clustering algorithms. Subsequently, a new large-scale clustering method based on a joint optimization of feature extraction and classification is proposed and applied to various databases, including two hyperspectral remote sensing images. The algorithm makes uses of a functional model (e.g., a neural network) for clustering which is trained by stochastic gradient descent. Results indicate that such a technique can easily scale to huge databases, can avoid the so-called out-of-sample problem, and can compete with or even outperform existing clustering algorithms on both artificial data and real remote sensing images. This is verified on small databases as well as very large problems. Résumé : Ce travail de recherche porte sur le développement et l'application de méthodes d'apprentissage dites non supervisées. Les applications visées par ces méthodes sont l'analyse de données forensiques et la classification d'images hyperspectrales en télédétection. Dans un premier temps, une méthodologie de classification non supervisée fondée sur l'optimisation symbolique d'une mesure de distance inter-échantillons est proposée. Cette mesure est obtenue en optimisant une fonction de coût reliée à la préservation de la structure de voisinage d'un point entre l'espace des variables initiales et l'espace des composantes principales. Cette méthode est appliquée à l'analyse de données forensiques et comparée à un éventail de méthodes déjà existantes. En second lieu, une méthode fondée sur une optimisation conjointe des tâches de sélection de variables et de classification est implémentée dans un réseau de neurones et appliquée à diverses bases de données, dont deux images hyperspectrales. Le réseau de neurones est entraîné à l'aide d'un algorithme de gradient stochastique, ce qui rend cette technique applicable à des images de très haute résolution. Les résultats de l'application de cette dernière montrent que l'utilisation d'une telle technique permet de classifier de très grandes bases de données sans difficulté et donne des résultats avantageusement comparables aux méthodes existantes.

Cardiac and vascular hypertrophy in Fabry disease: evidence for a new mechanism independent of blood pressure and glycosphingolipid deposition.

OBJECTIVE: Fabry disease is an X-linked disorder resulting from alpha-galactosidase A deficiency. The cardiovascular findings include left ventricular hypertrophy (LVH) and increased intima-media thickness of the common carotid artery (CCA IMT). The current study examined the possible correlation between these parameters. To corroborate these clinical findings in vitro, plasma from Fabry patients was tested for possible proliferative effect on rat vascular smooth muscle cells (vascular smooth muscle cell [VSMC]) and mouse neonatal cardiomyocytes. METHODS AND RESULTS: Thirty male and 38 female patients were enrolled. LVH was found in 60% of men and 39% of women. Increased CCA IMT was equally present in males and females. There was a strong positive correlation between LV mass and CCA IMT (r2=0.27; P<0.0001). VSMC and neonatal cardiomyocyte proliferative response in vitro correlated with CCA IMT (r2=0.39; P<0.0004) and LV mass index (r2=0.19; P=0.028), respectively. CONCLUSIONS: LVH and CCA IMT occur concomitantly in Fabry suggesting common pathogenesis. The underlying cause may be a circulating growth-promoting factor whose presence has been confirmed in vitro.

New Iowa materials : a supplement to the third edition of Iowa and some Iowans, 1990

This bibliography is an update to the publication Iowa and Some Iowans, 1988. It includes literature and history with elementary and secondary interests.

De la bellesa ornamental a la bellesa comercial : De les relacions entre art i publicitat en l'obra cartellista d'Alphonse Mucha

En aquest treball s'investiga, a través de l'obra d'Alphonse Mucha com autor paradigmàtic de l'Art Nouveau en el context geogràfic del París de finals del segle XIX i principis del XX, la possible influència de la publicitat en la difusió i consolidació de les creences i valors dels grups socials més ben posicionats, i molt especialment en el cas de la dona, ja que l'obra de Mucha està centrada bàsicament en la figura femenina. Així mateix, també es busca esbrinar la intervenció d'aquests grups en la creació d'un camp artístic que alhora faria possible la contemplació estètica de l'anunci comercial, i que esdevindria un element més d'aquesta moderna cultura, capitalista i urbana, de grans transformacions i avenços tecnològics, i que des d'una òptica crítica com la de l'Escola de Frankfurt, estaria 'cosificant' l'individu en nom de la ciència i el progrés, des del moment en què aquest quedaria condemnat al consum il·limitat per tal de satisfer un desig insaciable: l'afany d'emulació de la classe ociosa, o dit d'altra manera, d'aquella que el domina.

Plagiarism meets paraphrasing: insights for the new generation in automatic plagiarism detection

Although paraphrasing is the linguistic mechanism underlying many plagiarism cases, little attention has been paid to its analysis in the framework of automatic plagiarism detection. Therefore, state-of-the-art plagiarism detectors find it difficult to detect cases of paraphrase plagiarism. In this article, we analyse the relationship between paraphrasing and plagiarism, paying special attention to which paraphrase phenomena underlie acts of plagiarism and which of them are detected by plagiarism detection systems. With this aim in mind, we created the P4P corpus, a new resource which uses a paraphrase typology to annotate a subset of the PAN-PC-10 corpus for automatic plagiarism detection. The results of the Second International Competition on Plagiarism Detection were analysed in the light of this annotation. The presented experiments show that (i) more complex paraphrase phenomena and a high density of paraphrase mechanisms make plagiarism detection more difficult, (ii) lexical substitutions are the paraphrase mechanisms used the most when plagiarising, and (iii) paraphrase mechanisms tend to shorten the plagiarized text. For the first time, the paraphrase mechanisms behind plagiarism have been analysed, providing critical insights for the improvement of automatic plagiarism detection systems.

La autoreferencialidad en el arte. El metalenguaje en el medio digital. L'autoreferencialitat en l'art. El metallenguatge en el mitjà digital.

Este artículo analiza las obras de algunos artistas digitales que utilizan los recursos propios del medio digital como mensaje central. Nuestra intención será exponer las múltiples formas de utilizar, desarrollar y manipular el metalenguaje. Primeramente, con la intención de comprender estas obras, se elabora un contexto en que se plantearán unos precedentes, para ver, en una segunda parte, cómo el metalenguaje en nuestra actualidad artística y digital vuelve a ser muy utilizado, de manera que llega a convertirse en una metodología característica del arte contemporáneo y de los medios digitales. Palabras clave aleatoriedad, simulación, glitch, net art, interfaz, reactividad, interactividad

Active deformation fields: dense deformation field estimation for atlas-based segmentation using the active contour framework.

This paper presents a new and original variational framework for atlas-based segmentation. The proposed framework integrates both the active contour framework, and the dense deformation fields of optical flow framework. This framework is quite general and encompasses many of the state-of-the-art atlas-based segmentation methods. It also allows to perform the registration of atlas and target images based on only selected structures of interest. The versatility and potentiality of the proposed framework are demonstrated by presenting three diverse applications: In the first application, we show how the proposed framework can be used to simulate the growth of inconsistent structures like a tumor in an atlas. In the second application, we estimate the position of nonvisible brain structures based on the surrounding structures and validate the results by comparing with other methods. In the final application, we present the segmentation of lymph nodes in the Head and Neck CT images, and demonstrate how multiple registration forces can be used in this framework in an hierarchical manner.

Detection of metabolite induction in fungal co-cultures on solid media by high-throughput differential ultra-high pressure liquid chromatography-time-of-flight mass spectrometry fingerprinting.

Access to new biological sources is a key element of natural product research. A particularly large number of biologically active molecules have been found to originate from microorganisms. Very recently, the use of fungal co-culture to activate the silent genes involved in metabolite biosynthesis was found to be a successful method for the induction of new compounds. However, the detection and identification of the induced metabolites in the confrontation zone where fungi interact remain very challenging. To tackle this issue, a high-throughput UHPLC-TOF-MS-based metabolomic approach has been developed for the screening of fungal co-cultures in solid media at the petri dish level. The metabolites that were overexpressed because of fungal interactions were highlighted by comparing the LC-MS data obtained from the co-cultures and their corresponding mono-cultures. This comparison was achieved by subjecting automatically generated peak lists to statistical treatments. This strategy has been applied to more than 600 co-culture experiments that mainly involved fungal strains from the Fusarium genera, although experiments were also completed with a selection of several other filamentous fungi. This strategy was found to provide satisfactory repeatability and was used to detect the biomarkers of fungal induction in a large panel of filamentous fungi. This study demonstrates that co-culture results in consistent induction of potentially new metabolites.

Entropic transport in confined media: a challenge for computational studies in biological and soft-matter systems

Transport in small-scale biological and soft-matter systems typically occurs under confinement conditions in which particles proceed through obstacles and irregularities of the boundaries that may significantly alter their trajectories. A transport model that assimilates the confinement to the presence of entropic barriers provides an efficient approach to quantify its effect on the particle current and the diffusion coefficient. We review the main peculiarities of entropic transport and treat two cases in which confinement effects play a crucial role, with the appearance of emergent properties. The presence of entropic barriers modifies the mean first-passage time distribution and therefore plays a very important role in ion transport through micro- and nano-channels. The functionality of molecular motors, modeled as Brownian ratchets, is strongly affected when the motor proceeds in a confined medium that may constitute another source of rectification. The interplay between ratchet and entropic rectification gives rise to a wide variety of dynamical behaviors, not observed when the Brownian motor proceeds in an unbounded medium. Entropic transport offers new venues of transport control and particle manipulation and new ways to engineer more efficient devices for transport at the nanoscale.

Propuestas para afrontar los retos del European Credit Transfer System (ECTS) en asignaturas universitarias de redacción periodística

Los retos del nuevo paradigma comunicativo, un cruce histórico lleno de posibilidades pero también de riesgos y límites, obligan auna redefinición en la educación superior de los futuros profesionales de la comunicación.Este trabajo, realizado a partir de las experiencias de dos profesores universitarios, propone la introducción del hipertexto en ladocencia universitaria en el campo del periodismo y analiza las ventajas que esta nuevaherramienta supone en la formación de los periodistas del siglo XXI. En esta investigación los autores proponen el hipertexto como metodologíade trabajo en el aula, suponiendoque el hipertexto tiene que ser concebido,como se verá en esta investigación, como ellocus en el que deben converger la teoría y lapráctica, donde tiene que explotarse la capacidadcrítica y la creatividad del alumno y en elque debe potenciarse la interactividad profesor-alumno y alumno-alumno.