Hierarchical graph maps for visualization of collaborative recommender systems

In this paper we provide a method that allows the visualization of similarity relationships present between items of collaborative filtering recommender systems, as well as the relative importance of each of these. The objective is to offer visual representations of the recommender system?s set of items and of their relationships; these graphs show us where the most representative information can be found and which items are rated in a more similar way by the recommender system?s community of users. The visual representations achieved take the shape of phylogenetic trees, displaying the numerical similarity and the reliability between each pair of items considered to be similar. As a case study we provide the results obtained using the public database Movielens 1M, which contains 3900 movies.

A new label fusion method using graph cuts: application to hippocampus segmentation

The aim of this paper is to develop a probabilistic modeling framework for the segmentation of structures of interest from a collection of atlases. Given a subset of registered atlases into the target image for a particular Region of Interest (ROI), a statistical model of appearance and shape is computed for fusing the labels. Segmentations are obtained by minimizing an energy function associated with the proposed model, using a graph-cut technique. We test different label fusion methods on publicly available MR images of human brains.

Heterogeneous architecture for the optimization of large-scale graph processing data centres

Con el auge del Cloud Computing, las aplicaciones de proceso de datos han sufrido un incremento de demanda, y por ello ha cobrado importancia lograr m�ás eficiencia en los Centros de Proceso de datos. El objetivo de este trabajo es la obtenci�ón de herramientas que permitan analizar la viabilidad y rentabilidad de diseñar Centros de Datos especializados para procesamiento de datos, con una arquitectura, sistemas de refrigeraci�ón, etc. adaptados. Algunas aplicaciones de procesamiento de datos se benefician de las arquitecturas software, mientras que en otras puede ser m�ás eficiente un procesamiento con arquitectura hardware. Debido a que ya hay software con muy buenos resultados en el procesamiento de grafos, como el sistema XPregel, en este proyecto se realizará una arquitectura hardware en VHDL, implementando el algoritmo PageRank de Google de forma escalable. Se ha escogido este algoritmo ya que podr��á ser m�ás eficiente en arquitectura hardware, debido a sus características concretas que se indicaráan m�ás adelante. PageRank sirve para ordenar las p�áginas por su relevancia en la web, utilizando para ello la teorí��a de grafos, siendo cada página web un vértice de un grafo; y los enlaces entre páginas, las aristas del citado grafo. En este proyecto, primero se realizará un an�álisis del estado de la técnica. Se supone que la implementaci�ón en XPregel, un sistema de procesamiento de grafos, es una de las m�ás eficientes. Por ello se estudiará esta �ultima implementaci�ón. Sin embargo, debido a que Xpregel procesa, en general, algoritmos que trabajan con grafos; no tiene en cuenta ciertas caracterí��sticas del algoritmo PageRank, por lo que la implementaci�on no es �optima. Esto es debido a que en PageRank, almacenar todos los datos que manda un mismo v�értice es un gasto innecesario de memoria ya que todos los mensajes que manda un vértice son iguales entre sí e iguales a su PageRank. Se realizará el diseño en VHDL teniendo en cuenta esta caracter��ística del citado algoritmo,evitando almacenar varias veces los mensajes que son iguales. Se ha elegido implementar PageRank en VHDL porque actualmente las arquitecturas de los sistemas operativos no escalan adecuadamente. Se busca evaluar si con otra arquitectura se obtienen mejores resultados. Se realizará un diseño partiendo de cero, utilizando la memoria ROM de IPcore de Xillinx (Software de desarrollo en VHDL), generada autom�áticamente. Se considera hacer cuatro tipos de módulos para que as�� el procesamiento se pueda hacer en paralelo. Se simplificar�á la estructura de XPregel con el fin de intentar aprovechar la particularidad de PageRank mencionada, que hace que XPregel no le saque el m�aximo partido. Despu�és se escribirá el c�ódigo, realizando una estructura escalable, ya que en la computación intervienen millones de páginas web. A continuación, se sintetizar�á y se probará el código en una FPGA. El �ultimo paso será una evaluaci�ón de la implementaci�ón, y de posibles mejoras en cuanto al consumo.

Pedestrian Motion Prediction: A Graph Based Approach

A novel pedestrian motion prediction technique is presented in this paper. Its main achievement regards to none previous observation, any knowledge of pedestrian trajectories nor the existence of possible destinations is required; hence making it useful for autonomous surveillance applications. Prediction only requires initial position of the pedestrian and a 2D representation of the scenario as occupancy grid. First, it uses the Fast Marching Method (FMM) to calculate the pedestrian arrival time for each position in the map and then, the likelihood that the pedestrian reaches those positions is estimated. The technique has been tested with synthetic and real scenarios. In all cases, accurate probability maps as well as their representative graphs were obtained with low computational cost.

The case for unification.

I investigate the issue of whether the various subclasses of radio-loud galaxies are intrinsically the same but have been classified differently mainly due to their being viewed from different directions. Evidence for the two key elements of this popular version of the "unified scheme (US)," relativistic jets and nuclear tori, is updated. The case for the torus opening angle increasing with the radio luminosity of the active galactic nucleus (AGN) is freshly argued. Radio-loud AGN are particularly suited for testing the US, since their structures and polarization properties on different scales, as well as their overall radio sizes, provide useful statistical indicators of the relative orientations of their various subclasses. I summarize recent attempts to bring under a single conceptual framework the USs developed for radio-moderate [Fanaroff-Riley type I (FRI)] and radio-powerful (FRII) AGN. By focusing on FRII radio sources, I critically examine the recent claims of conflict with the US, based on the statistics of radio-size measurements for large, presumably orientation-independent, samples with essentially complete optical identifications. Possible ways of reconciling these results, and also the ones based on very-long-baseline radio interferometry polarimetric observations, with the US are pointed out. By incorporating a highly plausible temporal evolution of radio source properties into the US, I outline a scenario that allows the median linear size of quasars to approach, or even exceed, that of radio galaxies, as samples with decreasing radio luminosity are observed. Thus, even though a number of issues remain to be fully resolved, the scope of unified models continues to expand.

Difficulties with unification.

The difficulties perceived in the orientation-based unified scheme models, when confronted with the observational data, are pointed out. It is shown that in meter-wavelength selected samples, which presumably are largely free of an orientation bias, the observed numbers of quasars versus radio galaxies are not in accordance with the expectations of the unified scheme models. The observed number ratios seem to depend heavily on the redshift, fluxdensity, or radio luminosity levels of the selected sample. This cannot be explained within the simple orientation-based unified scheme with a fixed average value of the half-opening angle (c approximately 45 degrees ) for the obscuring torus that supposedly surrounds the nuclear optical continuum and the broad-line regions. Further, the large differences seen between radio galaxies and quasars in their size distributions in the luminosity-redshift plane could not be accommodated even if I were to postulate some suitable cosmological evolution of the opening angle of the torus. Some further implications of these observational results for the recently proposed modified versions of the unified scheme model are pointed out.

Unification and large-scale structure.

The hypothesis of relativistic flow on parsec scales, coupled with the symmetrical (and therefore subrelativistic) outer structure of extended radio sources, requires that jets decelerate on scales observable with the Very Large Array. The consequences of this idea for the appearances of FRI and FRII radio sources are explored.

Optical evidence for the unification of active galactic nuclei and quasi-stellar objects.

There is a variety of optical evidence for some unification of different types of active galactic nuclei and quasi-stellar objects (QSOs). The case is very strong for the unification of at least some Seyfert galaxies, where polarization data show that the type assigned to the Seyfert galaxy must depend on viewing direction. It has been proposed that Fanaroff-Riley type 2 (FR2) radio galaxies are quasars seen in a direction from which the quasar is obscured, and there is some limited direct evidence for this picture. The broad absorption line QSOs may be normal QSOs seen from a special direction. Some of the sources observed to have high luminosities in the far infrared could be obscured QSOs and active nuclei. Mergers and interactions are likely to play an important role in nuclear activity, and active galaxies and QSOs could change their apparent types through these encounters followed by subsequent evolution.

Electroencephalogram based Causality Graph Analysis in Behavior Tasks of Parkinson’s Disease Patients

Electroencephalographic (EEG) signals of the human brains represent electrical activities for a number of channels recorded over a the scalp. The main purpose of this thesis is to investigate the interactions and causality of different parts of a brain using EEG signals recorded during a performance subjects of verbal fluency tasks. Subjects who have Parkinson's Disease (PD) have difficulties with mental tasks, such as switching between one behavior task and another. The behavior tasks include phonemic fluency, semantic fluency, category semantic fluency and reading fluency. This method uses verbal generation skills, activating different Broca's areas of the Brodmann's areas (BA44 and BA45). Advanced signal processing techniques are used in order to determine the activated frequency bands in the granger causality for verbal fluency tasks. The graph learning technique for channel strength is used to characterize the complex graph of Granger causality. Also, the support vector machine (SVM) method is used for training a classifier between two subjects with PD and two healthy controls. Neural data from the study was recorded at the Colorado Neurological Institute (CNI). The study reveals significant difference between PD subjects and healthy controls in terms of brain connectivities in the Broca's Area BA44 and BA45 corresponding to EEG electrodes. The results in this thesis also demonstrate the possibility to classify based on the flow of information and causality in the brain of verbal fluency tasks. These methods have the potential to be applied in the future to identify pathological information flow and causality of neurological diseases.

Topological SLAM using a graph-matching based method on omnidirectional images

Comunicación presentada en el XI Workshop of Physical Agents, Valencia, 9-10 septiembre 2010.