Search

Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2023, Director: Ignasi Cos Aguilera, [en] The aim of this project is to establish a methodology to quantify the extent to which different brainwave signatures vary in their past data retention, and to determine how other factors interact with these variations, using previously obtained EEG recordings. To achieve this, we quantified the amount of past values that strongly influence future values for each brainwave signature throughout different EEG time series. Specifically, we calculated the number of lags required for a univariate autoregressive computational model to predict a set of brainwave time-series with an error (RMSE) below a preset threshold. In this fashion, we could establish that a similar number of lags were required based on the brainwave signatures (alpha, beta, gamma, or unfiltered) throughout the different conditions of activities and the different EEG sets, with the number of lags required being around 3, 4, and 6 for alpha, beta, and gamma brainwaves, respectively, when trying to achieve a minimum RMSE value of 0.001. This covariation is displayed again when using a different sets of threshold RMSE values, with gamma consistently having a greater dependency to past data, and alpha a lesser one. Our results indicate that brainwave signatures that are more related to active states can use past data for a longer period of time than brainwave signatures related to relaxed states. Furthermore, they suggest that active-state brainwaves show a more dilated time perception than their relaxed counterparts. In future studies, this methodology may help to establish a technique to objectively analyze time perception variation through EEG readings.

This article was written as a basis for a talk given in 2016 in the IQTCUB. In 2021 it was submitted for publication in The Journal of Chemical Education. The Editor-in-Chief Thomas Holme rejected it with arguments such as “While we are interested in articles that provide content and motivations for including such content in the curriculum, we are also aware that essentially every course in the chemistry curriculum is already overcrowded with content” and “there are no citations to literature in chemistry education”. In 2023 it has been published, after a minor revision, under a CC-by Creative Commons license at the digital repository of the Universitat de Barcelona (http://hdl.handle.net/2445/197800)., Density functional theory (DFT) plays a crucial role in computational chemistry, but its introduction in the chemistry curricula poses a challenge to lecturers when it comes to laying down its foundation without using a complex mathematical formalism and to establishing its scope and limitations. This article aims at presenting a simple and clear derivation of the theory that shows up its very general character. It is based on the constrained-search approach, that bears a close parallelism with the wave-function variational theorem, a familiar tool in any quantum chemistry introduction. Notwithstanding its simplicity, this approach shows appealing features, such as providing an explicit connection of the ground-state density with the electronic hamiltonian and an expression for the universal functional appearing in the definition of the energy functional. These lead to a more tangible insight into the theory than the original derivation by Hohenberg and Kohn, which only guarantees the existence of that connection and this functional. On the other hand, the approach highlights that the theory may, in principle, treat open and closed shell ground states on an equal footing, without the need to impose any restriction about the state spin. An interesting result is recalled that guarantees that a Kohn-Sham scheme may always be set up, no matter the open or closed shell character of the ground state nor its degree of electron correlation.

Treballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona. Curs: 2022-2023. Tutor: Sergio Escalera Guerrero i Rubén Ballester Bautista, [en] The performance of a deep neural network (DNN) is dependent on its ability to generalize. This ability is often expressed in the difference in accuracy on a training and test set, or the generalization gap. Recent research has seen the use of topological data analysis to estimate this performance gap without the use of a test set. Here, persistent homology measures are derived from a weighted graph of neuron activation correlations (functional network graph). The resulting persistence diagram is vectorized by a number of statistical summaries and correlated with the generalization gap. However, the computational complexity of persistent homology calculations hinders the application to DNNs with a larger number of activations. Methods are needed to sample these activations without losing predictive power. This work assesses the effect of different sampling strategies on the resulting persistence diagrams and their summaries. These include (non-)stratified random sampling, three methods based on notions of neuron importance similar to those used in pruning, and one using $k$-means++. In line with previous research some of these strategies provide models for predicting the generalization gap with high accuracy. The investigations provide insight and open up new lines of research into the structure of the functional network activation graph.

Logistic regression as a modelling technique of rare binary dependent variables with much fewer events (ones) than non-events (zeros) tends to underestimate their probability of occurrence. The vast literature devoted to the prediction of rare binary data identifies several ways to improve predictive performance by making modifications to the likelihood estimation. We propose two weighting mechanisms for incorporation in a pseudo-likelihood estimation that improve the predictive capacity of rare binary responses in data collected in complex surveys. We multiply sampling weights by specific correctors that lead to lower root mean square errors for event observations in almost all deciles. A case study is discussed where this method is implemented to predict the probability of suffering a workplace accident in a logistic regression model that is estimated with data from a survey conducted in Ecuador.

Abstract This paper examines the role played by foreign ownership as a determinant of the environmental behaviour of Spanish manufacturing firms. Unlike previous studies, our analysis distinguishes between current expenditure on and investment in environmental protection as a way of examining potential differences in the firms’ short- and long-term strategies. The results show how foreign capital boosts expenditure and especially investment. We also highlight the relevance of accounting for heterogeneity at the firm level by focusing on certain structural characteristics that are key in the manufacturing sector, namely, the size of the firm and the sector of activity where it operates. Our results show that size has a positive effect on firms’ decisions about environmental protection. The empirical results also reveal that the beneficial impact of foreign investment is mainly channelled through small firms and it is most evident in industries which are subject to less environmental regulation. Our findings may have pertinent implications regarding the determinants of firms’ environmental behaviour and about the effectiveness of foreign participation as a mechanism for overcoming potential barriers in the adoption of environmental protection measures by small firms.

Treballs Finals de Grau de Belles Arts. Facultat de Belles Arts. Universitat de Barcelona, Curs: 2020-2021, Tutor: Morell, Jordi, III Premi a la Qualitat Lingüística dels Treballs Final de Grau de la Facultat de Belles Arts, [cat] El recull i tractament massiu de la informació dels usuaris en la seva interacció amb el món virtual dona lloc a cossos de dades, associats amb cadascuna de les seves identitats. Aquestes dades, però, romanen emmagatzemades en bases de dades lluny dels seus corresponents cossos físics. Aquest treball respon a una voluntat de conèixer el propi cos de dades, la seva extensió i el seu contingut. Mitjançant una investigació sobre aquest, acabo per quantificar-lo i transportar-lo des del medi virtual a un pla físic., [eng] The massive collection and processing of user information in their interaction with the virtual world gives rise to data bodies, associated with each of their identities. This data, however, remains stored in databases away from their corresponding physical bodies. This work responds to a desire to know my own data body, its extent and its content. Through research on it, I end up quantifying it and transporting it from the virtual environment to a physical plane.

Research centres, universities and public organisations create datasets that can be reused in research. Reusing data makes it possible to reproduce studies, generate new research questions and new knowledge, but it also gives rise to technical and ethical challenges. Part of these issues are repositories interoperability to accomplish FAIR principles or issues related to data privacy or anonymity. At the same time, funding institutions require that data management plans be submitted for grants, and research tends to be increasingly interdisciplinary. Interdisciplinarity may entail barriers for researchers to reuse data, such as a lack of skills to manipulate data, given that each discipline generates different types of data in different technical formats, often non-standardized. Additionally, the use of standards to validate data reuse and better metadata to find appropriate datasets seem necessary. This paper offers a review of the literature that addresses data reuse in terms of technical, ethical-related issues.

Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2021, Director: Oliver Díaz i Jairo Luzón, [en] This project shows the design and development of an architecture based on Docker technology with the aim of studying its performance through metrics and being able to generate the visualization of the data in real time to extract possible conclusions. To test the tool, data on the global incidence of COVID-19 have been used. The architecture consists of different parts that configure it and for its implementation the most current and innovative tools have been used. For the data collection layer, a Kafka bus has been used in streaming based on an API (Application Programming Interface) of data, for data storage MongoDB has been used as a database, for data processing and its subsequent Visualization of the metrics, Prometheus has been used together with Grafana, allowing a User Friendly visualization through graphs, and finally, for the visualization of the data, MongoDB Charts has been used, a tool that has allowed the visualization of the data through heat maps. It can be concluded that the architecture has been designed in such a way that it is capable of managing data from any type of data source in a generic and fully modular and scalable way.

En su trabajo «Las teorías económicas contempladas a través de una óptica nacional» (1999), los profesores Ernest Lluch y José Luís Cardoso examinaban una visión de la historia del pensamiento económico que, alternativa a la tradicional exé¬gesis, sostenía el interés de estudiar autores y corrientes rele¬vantes en contextos concretos que, por razones diversas, no habían entrado a formar parte del canon de la ciencia econó¬mica. Este estilo historiográfico contextual, alejado de la «re¬construcción racional» de Blaug, ha sido cultivado con éxito por historiadores interesados en entornos no productores, sino receptores de ideas económicas. Aunque su estatus científico ha sido cuestionado, esta línea ha proporcionado frutos de enorme interés, desvelando que los procesos de transmisión y recepción de ideas económicas, lejos de ser inocuos, revelan intereses, conflictos, agendas y otros elementos que determi¬nan las ideas económicas que circulan en un medio particular, su cronología y su influencia. Pablo Cervera realiza un trabajo magistral en esta dirección, compilando y presentando los escritos de un economista español lamentablemente poco conocido, mas cuya obra ocupa un lugar fundamental en la historia del pensamiento económico español del siglo xviii, Enrique Ramos. De biografía compleja y escurridiza, es el pri¬mer autor español en incluir la locución «economía política» en el título de una obra. La expresión no es inocente ni neutral.

MAP2XANES is an intuitive Jupyter notebook that automatizes the analysis of synchrotron X-ray fluorescence imaging and X-ray absorption spectroscopy for the characterization of complex and heterogeneous samples. The notebook uses basic modules and functions from Numpy, Scipy, Pandas, iPywidgets and Matplotlib libraries for a powerful data reduction process that, in just a few clicks, guides the user through the visualization of elemental maps, spaceresolved absorption spectra and their automatized analysis. In particular, by means of linear combination fit of the XANES spectra, the notebook determines the chemical species distribution in the sample under investigation. The direct output of the analysis process is the correlation between the different elemental distributions and the spatial localization of the chemical species detected. An application to mineralogy is thus presented, analyzing the Mn2+, Mn3+ and Mn4+ distribution in a mineral sample of hausmannite (Mn2+Mn2 3+O4), courtesy of the Museum of Natural Science of Barcelona. MAP2XANES is an intuitive Jupyter notebook that automatizes the analysis of synchrotron X-ray fluorescence imaging and X-ray absorption spectroscopy for the characterization of complex and heterogeneous samples. The notebook uses basic modules and functions from Numpy, Scipy, Pandas, iPywidgets and Matplotlib libraries for a powerful data reduction process that, in just a few clicks, guides the user through the visualization of elemental maps, spaceresolved absorption spectra and their automatized analysis. In particular, by means of linear combination fit of the XANES spectra, the notebook determines the chemical species distribution in the sample under investigation. The direct output of the analysis process is the correlation between the different elemental distributions and the spatial localization of the chemical species detected. An application to mineralogy is thus presented, analyzing the Mn2+, Mn3+ and Mn4+ distribution in a mineral sample of hausmannite (Mn2+Mn2 3+O4), courtesy of the Museum of Natural Science of Barcelona. MAP2XANES is an intuitive Jupyter notebook that automatizes the analysis of synchrotron X-ray fluorescence imaging and X-ray absorption spectroscopy for the characterization of complex and heterogeneous samples. The notebook uses basic modules and functions from Numpy, Scipy, Pandas, iPywidgets and Matplotlib libraries for a powerful data reduction process that, in just a few clicks, guides the user through the visualization of elemental maps, spaceresolved absorption spectra and their automatized analysis. In particular, by means of linear combination fit of the XANES spectra, the notebook determines the chemical species distribution in the sample under investigation. The direct output of the analysis process is the correlation between the different elemental distributions and the spatial localization of the chemical species detected. An application to mineralogy is thus presented, analyzing the Mn2+, Mn3+ and Mn4+ distribution in a mineral sample of hausmannite (Mn2+Mn2 3+O4), courtesy of the Museum of Natural Science of Barcelona.

A boosting-based machine learning algorithm is presented to model a binary response with large imbalance, i.e., a rare event. The new method (i) reduces the prediction error of the rare class, and (ii) approximates an econometric model that allows interpretability. RiskLogitboost regression includes a weighting mechanism that oversamples or undersamples observations according to their misclassification likelihood and a generalized least squares bias correction strategy to reduce the prediction error. An illustration using a real French third-party liability motor insurance data set is presented. The results show that RiskLogitboost regression improves the rate of detection of rare events compared to some boosting-based and tree-based algorithms and some existing methods designed to treat imbalanced responses.

Aquest document conté originàriament altre material i/o programari només consultable a la Biblioteca de Ciència i Tecnologia. El presente proyecto tiene como objetivo desarrollar una tecnología que permita codificar grandes cantidades de texto de manera automática para posteriormente ser visualizada y analizada mediante una aplicación diseñada en Qlikview. El motor de la investigación e implementación de este proyecto se ha encontrado en la incipiente presencia de tecnologías informáticas en los procesos de codificación para ciencias políticas. De esta manera, el programa creado tiene como objetivo automatizar un proceso que se desarrolla comúnmente de manera manual y, por ende, las ventajas de introducir técnicas informáticas son notablemente valiosas. Estas automatizaciones permiten ahorrar tanto en tiempo de codificación, como en recursos económicos o humanos. Se ha elaborado una revisión teórica y metodológica que han servido como instrumentos de estudio y mejora, con el firme propósito de reducir al máximo el margen de error y ofrecer un instrumento de calidad con salida de mercado real. El método de clasificación utilizado ha sido Bayes, y se ha implementado utilizando Matlab. Los resultados de la clasificación han llegado a índices del 99.2%. En la visualización y análisis mediante Qlikview se pueden modificar los parámetros referentes a partido político, año, categoría o región, con lo que se permite analizar numerosos aspectos relacionados con la distribución de las palabras repartidas entre las diferentes categorías y en el tiempo. The aim of the present project is to develop a technology capable of codifying a huge amount of text automatically in order to be analysed through a Qlikview application. The main reason for the investigation and implementing of this project has been found due to the new presence of information technology in codifying processes for political science. Thus, the program created seeks to automate a usually hand-made process, and the advantages of introducing these techniques are remarkably valuable. This automates allow to save time both in time and in economic or human resources. In this point, there has been a theoretic and methodological revision that worked out as study and development instruments, with the aim to reduce the margin of error and offering a quality tool with access to real market. The classification method used has been Bayes, and it has been implemented by using Matlab. The classification results have reached 99.2% success. In the visualization and analysis with Qlikview the values of political party, year, category or region can be modified allowing analyse numerous aspects related to the word distribution between the categories and through time. El present projecte té com objectiu desenvolupar una tecnologia que permeti codificar grans quantitats de text de forma automàtica per a posteriorment ser visualitzada i analitzada mitjançant una aplicació dissenyada en Qlikview. El motor de la investigació i implementació d'aquest projecte s'ha trobat en la incipient presència de tecnologies informàtiques als processos de codificació per a ciències polítiques. D'aquesta forma, el programa creat té com a objectiu automatitzar un procés que es desenvolupa comunament de forma manual i, per tant, els avantatges d'introduir tècniques informàtiques són molt valuoses. Aquestes automatitzacions permeten tant estalviar temps de codificació, com en recursos econòmics o humans. S'ha preparat una revisió teòrica i metodològica que ha funcionat com a instrument d'estudi i de millora, amb el ferma propòsit de reduir al màxim el marge d'error i oferir un instrument de qualitat amb sortides al mercat real. El mètode de classificació utilitzat ha estat Bayes, i s'ha implementat mitjançant Matlab. Els resultats de la classificació han arribat a índexs del 99.2%. A la visualització i anàlisis mitjançant Qlikview es poden modificar el paràmetres referents a partit polític, any, categoria o regió, amb el que es poden analitzar nombrosos aspectes relacionats amb la distribució de les paraules repartides entre les diferents categories i en el temps.

Treballs Finals del Màster de Planificació Territorial i Gestió Ambiental, Facultat de Geografia i Història, Universitat de Barcelona, Curs: 2020-2021, Tutor: Jaime Llorca Ballester, La problemática mundial acerca de la calidad del aire ha adquirido relevancia en la última década debido a recientes hallazgos sobre los efectos en la salud humana. Efectivamente, el último informe de la Organización Mundial de la Salud (OMS) indicó que alrededor de siete millones de personas mueren al año a causa de los altos niveles de exposición a contaminantes atmosféricos criterio (OECD, 2017). Específicamente, la alta tasa de morbilidad y mortalidad está relacionada con la formación de material particulado fino (PM2.5), ingresando por las vías respiratorias o de manera oral al organismo humano y ocasionando enfermedades en los pulmones, bronquios, alvéolos e incluso en algunos casos genera efectos en el sistema circulatorio (Gualtieri et al., 2011). Por lo expuesto anteriormente, en este caso de estudio se determinan las concentraciones de material particulado (PM2.5) a partir de los registros de Subsistema de información de la calidad del aire (SISAIRE). Asimismo, de acuerdo con los registros obtenidos y sus datos meteorológicos asociados a su ubicación, se elabora un modelo de dispersión de contaminantes por medio del software Screen View 4.0 que hace parte del conjunto de programas de “Enviromental Lakes”, además, se realiza un modelo geoespacial de los contaminantes mediante el software ArcGIS 10.5. Después, teniendo en cuenta los resultados de dichos modelos sumado a los reportes de salud pública obtenidos de los Registros Individuales de la Prestación de Salud (RIPS), se establece una correlación estadística entre las enfermedades respiratorias de la población de Sogamoso y los niveles de inmisión del municipio. Finalmente, con base en los resultados obtenidos se definen soluciones para asegurar la calidad del aire y la salud humana.

El següent treball s’ha realitzat amb l’objectiu d’implementar una eina que permeti i faciliti l’obtenció de l’arbre de discussió que es genera en els comentaris d’un vídeo de YoutTube. Després d’un primer anàlisi sobre la informació que ens ofereix la seva API, s’ha detectat que YouTube aplica una simplificació a l’hora d’emmagatzemar els comentaris dels usuaris fent que es perdi informació rellevant per tal de crear l’arbre de discussió. Quan YouTube guarda els comentaris pertanyents a un fil, assigna com a pare de tots aquests el principal. Fet que, si hi ha algun comentari que contesta a un altre se li assigni com a pare el principal en canvis del que contesta. Així doncs, s’ha pensat a implementar una solució que pugui estar a l’abast de tothom en cas que algú es trobi amb el mateix problema, i és per això que s’ha decidit desenvolupar una llibreria. Aquesta s’ha programat amb el Llenguatge de Programació Python i posteriorment s’ha publicat a PyPi org. Per posar solució a aquest conflicte de falta d’informació a l’hora de crear l’arbre de discussió, s’ha dissenyat un algoritme automàtic mitjan¸cant el qual es realitzarà un rànquing dels candidats a ser el pare d’un comentari i s’escollirà aquell més rellevant.

[eng] Complex systems can be understood as an entity composed by a large number of interactive elements whose emergent global behaviour cannot be derived from the local laws characterizing their constituents. The observables characterizing these systems can be observed at different scales and they often exhibit interesting properties such as lack of characteristic scales and self-similarity. In this context, power-law type functions take an important role in the description of these observables. The presence of power-law functions resembles to the situation of thermodynamic quantities close to a critical point in equilibrium critical phenomena. Different complex systems can be grouped into the same universality class when the power-law functions characterizing their observables have the same exponents. The response of some complex systems proceeds by the so called avalanche process: a collective response of the system characterized by following an intermittent dynamics, with sudden bursts of activity separated by periods of silence. This kind of out-of-equilibrium systems can be found in different disciplines such as seismology, astrophysics, ecology, finance or epidemiology, just to mention a few of them. Avalanches are characterized by a set of observables such as the size, the duration or the energy. When avalanche observables exhibit lack of characteristic scales, their probability distributions can be statistically modelled by power-law-type distributions. Avalanche criticality occurs when avalanche observables can be characterized by this kind of distributions. In this sense, the concepts of criticality and universality, which are well defined in equilibrium phenomena, can be also extended for the probability distributions describing avalanche observables in out-of-equilibrium systems. The main goal of this PhD thesis relies on providing robust statistical methods in order to characterize avalanche criticality and universality in empirical datasets. Due to limitations in data acquisition, empirical datasets often only cover a narrow range of observation, making it difficult to establish power-law behaviour unambiguously. With the aim of discussing the concepts of avalanche criticality and universality, two different systems are going to be considered: earthquakes and acoustic emission events generated during compression experiments of porous materials in the laboratory (labquakes). The techniques developed in this PhD thesis are mainly focused on the distribution of earthquake and labquake sizes, which is known as the Gutenberg-Richter law. However, the methods are much more general and can be applied to any other avalanche observable. The statistical techniques provided in this work can also be helpful for earthquake forecasting. Coulomb-stress theory has been used for years in seismology to understand how earthquakes trigger each other. Earthquake models that relate earthquake rates and Coulomb stress after a main event, such as the rate-and-state model, assume that the magnitude distribution of earthquakes is not affected by the change in the Coulomb stress. Several statistical analyses are performed to test whether the distribution of magnitudes is sensitive to the sign of the Coulomb-stress increase. The use of advanced statistical techniques for the analysis of complex systems has been found to be necessary and very helpful in order to provide rigour to the empirical results, particularly, to those problems regarding hazard analysis., [cat] Els sistemes complexos es poden entendre com entitats compostes per un gran nombre d’elements en interacció on la seva resposta global i emergent no es pot derivar de les lleis particulars que caracteritzen cadascun dels seus constituents. Els observables que caracteritzen aquests sistemes es poden observar a diferents escales i, sovint, mostren propietats interessants tals com la manca d’escales característiques i autosimilitud. En aquest context, les funcions amb lleis de potència prenen un paper important en la descripció d’aquests observables. La presència de lleis de potència s’assimila a la situació dels fenòmens crítics en equilibri, on algunes quantitats termodinàmiques mostren un comportament funcional similar prop d’un punt crític. Diferents sistemes complexos es poden agrupar en la mateixa classe d’universalitat quan les funcions de lleis de potència que caracteritzen els seus observables tenen els mateixos exponents. Quan són conduïts externament, la resposta d’alguns sistemes complexos segueix el que s’anomonena un procès d’allaus: una resposta col·lectiva del sistema caracteritzada per seguir una dinàmica intermitent, amb sobtats increments d’activitat separats per períodes de silenci. Aquesta mena de sistemes fora de l’equilibri es poden trobar en diferents disciplines tals com la sismologia, astrofísica, ecologia, epidemologia o finances, per mencionar alguns. Les allaus estan caracteritzades per un conjunt d’observables tals com la mida, l’energia o la durada. Quan aquests observables mostren una manca d’escales característiques, les seves distribucions de probabilitat es poden modelitzar estadísticament per distribucions de lleis de potència. S’anomenen allaus crítiques aquelles en que els seus observables es poden caracteritzar per aquestes distribucions. En aquest sentit, els conceptes de criticalitat i universalitat, els quals estan ben definits per fenòmens en equilibri, es poden extendre per les distribucions de probabilitat que descriuen els observables de les allaus en sistemes fora de l’equilibri. L’objectiu principal d’aquesta tesi doctoral és proporcionar mètodes estadístics robusts per tal de caracteritzar la criticalitat i la universalitat en allaus corresponents a dades empíriques. Degut a les limitacions en l’adquisició de dades, les dades empíriques sovint cobreixen un rang petit d’observació, dificultant que es pugui establir un determinat comportament en forma de llei de potència de manera inequívoca. Amb l’objectiu de discutir els conceptes de criticalitat i universalitat en allaus, es consideraran dos sistemes diferents: els terratrèmols i els esdeveniments d’emissió acústica que es generen durant experiments de compressió de materials porosos al laboratori (labquakes). Les tècniques desenvolupades en aquesta tesi doctoral estan enfocades principalment a la distribució de la mida dels terratrèmols i labquakes, altrament coneguda com a llei de Gutenberg-Richter. No obstant, aquests mètodes són molt més generals i es poden aplicar a qualsevol observable de les allaus. Les tècniques estadístistiques proporcionades en aquest treball poden també ajudar al pronòstic de terratrèmols. Durant anys, la teoria d’esforços de Coulomb s’ha utilitzat en sismologia per tal d’entendre com els terratrèmols desencadenen l’ocurrència d’altres de nous. Els models de terratrèmols que relacionen la taxa d’ocurrència de rèpliques i l’esforç de Coulomb després d’un gran esdeveniment, assumeixen que la distribució de la mida dels terratrèmols no està afectada pel canvi en l’esforç de Coulomb. Diverses anàlisi estadístiques s’aplicaran per tal de comprovar si la distribució de magnituds és sensible al signe de l’esforç de Coulomb. S’ha provat que l’ús de tècniques estadístiques avançades en l’anàlisi de sistemes complexos és útil i necessari per tal d’aportar rigor als resultats empírics i, en particular, a problemes d’anàlisi de riscos.

Public administration scholarship reflects a multidisciplinary field in which many theoretical perspectives coexist. However, one of the dark sides of such theoretical pluralism is methodological fragmentation. It may be hard to assess the research quality and to engage with the findings from studies employing different methodologies, thus limiting meaningful conversations. Moreover, the constant race across social sciences to make methodologies more sophisticated may exacerbate the separation between academic and practitioner audiences. To counterbalance these two trends, this article aims at increasing methodological intelligibility in our field. It does so starting from the idea that each methodology entails choices in the conventional phases of research design, data collection, and data analysis, and that these choices must be reported. The paper nails down and exemplifies such reporting needs for five selected methodologies: survey studies, quantitative experimental studies, quantitative observational studies, qualitative case studies and ethnographies. Based on their discussion and comparison, the paper offers a framework composed by functional equivalents, that is to say, the common denominator among methodological reporting needs. Methodological choices that need reporting include the rationale for the selection of a methodology, delimitation of the study, the research instrument, data processing and ethical clearance. Increasing methodological reporting would facilitate dialogues among different methodological communities, and with practitioner readers. All of which would also promote field building in the scholarship of public administration.

We investigate the extent and evolution of the links between energy markets using a broad data set consisting of a total of 17 series of prices for commodities such as electricity, natural gas, coal, oil and carbon. The results shed light on a number of relevant issues such as the volatility spillover effect in energy markets (within and across sectors) and the identification of those markets that are exporters (importers) of volatility to (from) other markets, as well as evidence of the time-varying nature of these effects. The main conclusions are: (i) the most integrated European electricity markets appear to be those of Germany, France and the Netherlands; (ii) the Dutch Title Transfer Facility might be on the way to becoming the benchmark price for natural gas in Europe, and (iii) natural gas may be replacing crude oil as the global benchmark price for energy commodities.

En aquest treball, es plantejarà la construcció i estudi d'una plataforma que permeti realitzar l'anàlisi d'una nova metodologia per a la detecció d'imatges modificades, basada en la combinació de l'estrategia Divide And Conquer, utilitzan xarxes neuronals i les característiques de les imatges. In this project, the construction and study of a platform will be proposed to allow the analysis of a new methodology for the detection of modified images, based on the combination of the Divide And Conquer strategy, using neural networks trained with the properties of the image. En este proyecto se planteará la construcción y estudio de una plataforma que permita la realización y análisis de una nueva metodología para la detección de imágenes modificadas, basada en la combinación de la estrategia Divide And Conquer, utilizando redes neuronales y las propiedades de las imágenes.

Pay-how-you-drive (PHYD) or usage-based (UB) systems for automobile insurance provide actuaries with behavioural risk factors, such as the time of the day, average speeds and other driving habits. These data are collected while the contract is in force with the help of telematic devices installed in the vehicle. They thus fall in the category of a posteriori information that becomes available after contract initiation. For this reason, they must be included in the actuarial pricing by means of credibility updating mechanisms instead of being incorporated in the score as ordinary a priori observable features. This paper proposes the use of multivariate mixed models to describe the joint dynamics of telematics data and claim frequencies. Future premiums, incorporating past experience can then be determined using the predictive distribution of claim characteristics given past history. This approach allows the actuary to deal with the variety of situations encountered in insurance practice, ranging from new drivers without telematics record to contracts with different seniority and drivers using their vehicle to different extent, generating varied volumes of telematics data., SCOPUS: ar.j, info:eu-repo/semantics/published

Universities are organizational structures with individual activity mixes or strategies that lead to different performance levels by mission. Evaluation techniques based on performance indicators or rankings risk rewarding just a specific type of university and undermining university diversification: They usually introduce homogenizing pressures and risk displacing university objectives- neglecting their socio-economic contribution and focusing on succeeding on the evaluation system. In this study, we propose an alternative evaluation method that overcomes these limitations. We produce a multidimensional descriptive classification of universities into typologies, while analysing the relation between their institutional factors (characteristics) and their (technical) efficiency performance from a descriptive perspective. To do so we apply bootstrap data envelopment analysis (DEA) and multidimensional scaling analysis (MDS), performing a so-called DEAMDS analysis on data on the Spanish university system, and unlike previous studies, we include data on an important dimension of the third mission of universities (specifically knowledge transfer, KT) in their characterization. We identify six types of (homogeneous) universities. Results indicate that to be fairly efficient, universities may focus on teaching, KT, or overall efficiency but always have to fairly perform in research. Additionally, results confirm the relevance of the third mission as a source of institutional diversity in higher education. This approach could be used to address an alternative evaluation methodology for higher education institutions with formative purposes, evaluating universities according to their unique characteristics for the improvement of HE systems

The GEOTRACES Intermediate Data Product 2017 (IDP2017) is the second publicly available data product of the international GEOTRACES programme, and contains data measured and quality controlled before the end of 2016. The IDP2017 includes data from the Atlantic, Pacific, Arctic, Southern and Indian oceans, with about twice the data volume of the previous IDP2014. For the first time, the IDP2017 contains data for a large suite of biogeochemical parameters as well as aerosol and rain data characterising atmospheric trace element and isotope (TEI) sources. The TEI data in the IDP2017 are quality controlled by careful assessment of intercalibration results and multi-laboratory data comparisons at crossover stations. The IDP2017 consists of two parts: (1) a compilation of digital data for more than 450 TEIs as well as standard hydrographic parameters, and (2) the eGEOTRACES Electronic Atlas providing an on-line atlas that includes more than 590 section plots and 130 animated 3D scenes. The digital data are provided in several formats, including ASCII, Excel spreadsheet, netCDF, and Ocean Data View collection. Users can download the full data packages or make their own custom selections with a new on-line data extraction service. In addition to the actual data values, the IDP2017 also contains data quality flags and 1-σ data error values where available. Quality flags and error values are useful for data filtering and for statistical analysis. Metadata about data originators, analytical methods and original publications related to the data are linked in an easily accessible way. The eGEOTRACES Electronic Atlas is the visual representation of the IDP2017 as section plots and rotating 3D scenes. The basin-wide 3D scenes combine data from many cruises and provide quick overviews of large-scale tracer distributions. These 3D scenes provide geographical and bathymetric context that is crucial for the interpretation and assessment of tracer plumes near ocean margins or along ridges. The IDP2017 is the result of a truly international effort involving 326 researchers from 25 countries. This publication provides the critical reference for unpublished data, as well as for studies that make use of a large cross-section of data from the IDP2017. This article is part of a special issue entitled: Conway GEOTRACES - edited by Tim M. Conway, Tristan Horner, Yves Plancherel, and Aridane G. González., National Science Foundation (U.S.) (Grant OCE-0608600), National Science Foundation (U.S.) (Grant OCE0938349), National Science Foundation (U.S.) (Grant OCE-1243377), National Science Foundation (U.S.) (Grant OCE-1546580)

Context. The second Gaia data release is based on 22 months of mission data with an average of 0.9 billion individual CCD observations per day. A data volume of this size and granularity requires a robust and reliable but still flexible system to achieve the demanding accuracy and precision constraints that Gaia is capable of delivering. Aims. We aim to describe the input data, the treatment of blue photometer/red photometer (BP/RP) low-resolution spectra required to produce the integrated GBP and GRP fluxes, the process used to establish the internal Gaia photometric system, and finally, the generation of the mean source photometry from the calibrated epoch data for Gaia DR2. Methods. The internal Gaia photometric system was initialised using an iterative process that is solely based on Gaia data. A set of calibrations was derived for the entire Gaia DR2 baseline and then used to produce the final mean source photometry. The photometric catalogue contains 2.5 billion sources comprised of three different grades depending on the availability of colour information and the procedure used to calibrate them: 1.5 billion gold, 144 million silver, and 0.9 billion bronze. These figures reflect the results of the photometric processing; the content of the data release will be different due to the validation and data quality filters applied during the catalogue preparation. The photometric processing pipeline, PhotPipe, implements all the processing and calibration workflows in terms of Map/Reduce jobs based on the Hadoop platform. This is the first example of a processing system for a large astrophysical survey project to make use of these technologies. Results. The improvements in the generation of the integrated G–band fluxes, in the attitude modelling, in the cross-matching, and and in the identification of spurious detections led to a much cleaner input stream for the photometric processing. This, combined with the improvements in the definition of the internal photometric system and calibration flow, produced high-quality photometry. Hadoop proved to be an excellent platform choice for the implementation of PhotPipe in terms of overall performance, scalability, downtime, and manpower required for operations and maintenance.

Treballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona, Any: 2018, Tutor: Mireia Ribera i Eloi Puertas i Prats, [en] This thesis concludes the Masters studies in Data Science of the University of Barcelona during the course 2017/2018. It contains a study of the escoltas catalans census, an non-profit organization, part of the scouts foundation, with the aim of obtaining insights that might be useful for enhacing their internal organization or assisting future decisions by adding information. The work done during this thesis includes different stages of the datascience pipeline, from data gathering and data enriching to the use of state of the art techniques for predicting future behaviours. The analysis performed on the obtained data resulted in two main outcomes: • Infographics. A visualization explaining data characteristics and some interesting facts found while performing the data analysis was created. In order to create this visualization, techniques learned during the course of the Presentation and Visualization subject were applied. • Analysis of implication. Aiming to have a better understanding of which factors determine the implication of the association members, a study of the length of stay and its correlated circumstances has been performed. Apart of those outcomes, the job done around data cleaning, formatting and enriching might be considered another interesting outcome of this project. To perform the analysis, open data has been leveraged to add the necessary information about the context and domain. To perform all done work regarding data acquisition, processing and analysis, different open source tools have been used. All the information and details about project outcomes and development are explained in the present document.

Context. The Gaia spacecraft of the European Space Agency (ESA) has been securing observations of solar system objects (SSOs) since the beginning of its operations. Data Release 2 (DR2) contains the observations of a selected sample of 14,099 SSOs. These asteroids have been already identified and have been numbered by the Minor Planet Center repository. Positions are provided for each Gaia observation at CCD level. As additional information, complementary to astrometry, the apparent brightness of SSOs in the unfiltered G band is also provided for selected observations. Aims. We explain the processing of SSO data, and describe the criteria we used to select the sample published in Gaia DR2. We then explore the data set to assess its quality. Methods. To exploit the main data product for the solar system in Gaia DR2, which is the epoch astrometry of asteroids, it is necessary to take into account the unusual properties of the uncertainty, as the position information is nearly one-dimensional. When this aspect is handled appropriately, an orbit fit can be obtained with post-fit residuals that are overall consistent with the a-priori error model that was used to define individual values of the astrometric uncertainty. The role of both random and systematic errors is described. The distribution of residuals allowed us to identify possible contaminants in the data set (such as stars). Photometry in the G band was compared to computed values from reference asteroid shapes and to the flux registered at the corresponding epochs by the red and blue photometers (RP and BP). Results. The overall astrometric performance is close to the expectations, with an optimal range of brightness G ∼ 12 - 17. In this range, the typical transit-level accuracy is well below 1 mas. For fainter asteroids, the growing photon noise deteriorates the performance. Asteroids brighter than G ∼ 12 are affected by a lower performance of the processing of their signals. The dramatic improvement brought by Gaia DR2 astrometry of SSOs is demonstrated by comparisons to the archive data and by preliminary tests on the detection of subtle non-gravitational effects., 0, info:eu-repo/semantics/published

This article describes the negation cue detection approach designed and built by UPC's team participating in NEGES 2018 Workshop on Negation in Spanish. The approach uses supervised CRFs as the base for training the model with several features engineered to tackle the task of negation cue detection in Spanish. The result is evaluated by the means of precision, recall, and F1 score in order to measure the performance of the approach. The approach was ranked in 1st position in the official testing results with average precision around 91%, average recall around 82%, and average F1 score around 86%. © 2018 CEUR-WS. All rights reserved.

Se trata de un compendio para poder iniciarse en el programa Iramuteq, que nos permite analizar texto de forma automática. Este software es de acceso libre y permite tener análisis globales sobre la información recogida tanto en cuestionarios o entrevistas como materiales de webs, diarios etc..

BACKGROUND: Reduction in the cost of genomic assays has generated large amounts of biomedical-related data. As a result, current studies perform multiple experiments in the same subjects. While Bioconductor's methods and classes implemented in different packages manage individual experiments, there is not a standard class to properly manage different omic datasets from the same subjects. In addition, most R/Bioconductor packages that have been designed to integrate and visualize biological data often use basic data structures with no clear general methods, such as subsetting or selecting samples. RESULTS: To cover this need, we have developed MultiDataSet, a new R class based on Bioconductor standards, designed to encapsulate multiple data sets. MultiDataSet deals with the usual difficulties of managing multiple and non-complete data sets while offering a simple and general way of subsetting features and selecting samples. We illustrate the use of MultiDataSet in three common situations: 1) performing integration analysis with third party packages; 2) creating new methods and functions for omic data integration; 3) encapsulating new unimplemented data from any biological experiment. CONCLUSIONS: MultiDataSet is a suitable class for data integration under R and Bioconductor framework.

Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2017, Director: Christos Verikoukis, This project creates an experimental platform, consists of moving the data storage and processing capabilities of the cloud at the edge of the network, in order to reduce latency times and ease the load of the network that communicates to cloud. The concept of the IoTWorld of the Centre Tecnològic de Telecomunicacions de Catalunya (CTTC) is detailed. Explain how the experimental platform will be carried out. It details the way in which IoT device data is simulated, so that it does not have to acquire real devices that generate them for the experimental platform. Explain the possibilities of moving the cloud to the edge, details the way in which the experimental platform will be tested and performed. Performance tests will be performed on a real Cloud, as well as on the infrastructure that has been developed to later compare them, so see if it is worth downloading the Cloud at the edge.

We introduce a new comparative response format, suitable for assessing personality and similar constructs. In this “graded-block” format, items measuring different constructs are first organized in blocks of 2 or more; then, pairs are formed from items within blocks. The pairs are presented one at a time, to enable respondents expressing the extent of preference for one item or the other using several graded categories. We model such data using confirmatory factor analysis (CFA) for ordinal outcomes. We derive Fisher information matrices for the graded pairs, and supply R code to enable computation of standard errors of trait scores. An empirical example illustrates the approach in low-stakes personality assessments and shows that similar results are obtained when using graded blocks of size 3 and a standard Likert format. However, graded-block designs may be superior when insufficient differentiation between items is expected (due to acquiescence, halo or social desirability).

[eng] The process of creating an efficacious malaria vaccine is complex due to the characteristics of the disease that are directly related to the responsible parasite. In the disease-vaccine interaction several aspects need to be taken into account to improve and understand the vaccine and for that reason different types of data need to be analyzed. Current assays technology allows analyzing several proteins simultaneously with a small blood volume. The combination of the medium throughput dataset of some assays and the small sample size of some malaria studies may hinder the use of classical statistical methods. In the context of low number of observations and medium or high number of variables the support vector machines (SVM) models are a powerful tool to analyze sparse data, i.e., data in which the number of predictors is larger or approximately equal to the number of observations, especially when handling binary outcomes. However, biomedical research often involves analysis of time-to-event outcomes. Several methods have been tested in the literature to deal with censored data into the SVM framework. Most of these methods are based on a support vector regression (SVR) approach and results found in the literature suggest no significant differences with Cox proportional hazards model and kernel Cox regression. Another perspective is a SVM for binary classification, however, almost no work has been done into this approach: only SVM learning using privileged information and SVM with uncertain classes have been described. This PhD thesis aims to propose alternative methods and extensions to the ones existing in the binary classification framework, specifically, proposing a conditional survival approach for weighting censored observations, a semi-supervised SVM with local invariances perspective and evaluating a weighted SVM model. Another important aspect in biomedical research is to identify the relevance of the variables in a model, i.e., which variables are important related to the response variable. In the SVM framework most of the work done is related to linear kernels, however, the main advantage of SVM is using non-linear kernels. This PhD thesis aims to propose three approaches based on the Recursive Feature Elimination (RFE) algorithm to rank variables based on non-linear SVM and SVM for survival analysis. Moreover, the proposed algorithms are focused on interpretation and visualization of each one the RFE iterations, allowing to identify relevant variables associated with the response variable and among predictor variables. After evaluating all proposed methods in a simulation study under several scenarios, a real dataset applying these methods has been analyzed: the Mal067 data aims to identify immune responses correlated with protection from malaria that were elicited by the malaria RTS,S vaccine and by natural immunity. All SVM for survival analysis methods have been implemented in R, since neither R packages nor R functions have been found., [cat] El procés de crear una vacuna eficaç contra la malària és complex degut a les característiques del paràsit responsable. Les tècniques de laboratori actuals permeten analitzar moltes proteïnes simultàniament amb molt poc volum de sang, això, juntament amb la poca grandària mostral de molts estudis de malària fa que els mètodes estadístics clàssics no siguin adequats. Les màquines de suport vectorial (SVM) són una eina molt potent per tractar aquest tipus de dades en el context de poques observacions i moltes variables, moltes vegades, però, la recerca està enfocada en variables resposta temps fins a esdeveniment. Gran part de la recerca feta a la literatura en aproximar els mètodes SVM a dades de supervivència està enfocada des de la perspectiva de SVM per regressió. Una altra perspectiva molt poc desenvolupada i avaluada és la de SVM per classificació binària. En aquesta tesi proposem extensions i mètodes alternatius basats en SVM per classificació binària, específicament, proposant una ponderació de les dades censurades basada en la supervivència condicionada, una perspectiva semi-supervisada de SVM amb invariàncies locals i l’avaluació de SVM ponderant les observacions censurades. Un aspecte important en la recerca biomèdica és la identificació de la rellevància de variables en el model, és a dir, quines variables són importants en relació a la variable resposta. En el context de SVM, gran part de la recerca està enfocada a kernels lineals, però el gran avantatge dels SVM és la possibilitat d’utilitzar kernels no lineals. En aquesta tesi proposem tres aproximacions basades en l’algoritme recursiu d’eliminació de característiques (RFE) per ordenar variables, des d’una perspectiva de kernels no lineals i SVM per l’anàlisi de supervivència. A més, els mètodes proposats permeten ser interpretats i visualitzats a cada iteració de l’algoritme RFE, permeten identificar la rellevància de les variables predictores amb respecte la variable resposta i l’associació entre variables predictores. Després d’avaluar tots els mètodes proposats per SVM amb dades censurades i rellevància de variables, mitjançant simulacions, s’han analitzat les dades reals de l’estudi Mal067 que estudia correlats de protecció contra la malària induïts per la vacuna RTS,S.

El projecte descrit en aquesta memòria consisteix en el desenvolupament d’una plataforma pel control ambiental d’una nau avícola amb l’objectiu de poder controlar el benestar animal i tenir una idea de l’evolució dels diferents paràmetres ambientals al llarg del temps. Per dur a terme aquest treball s’ha realitzat diversos estudis de mercat per definir els objectius principals on es volia arribar. També s’ha realitzat un estudi de les diferents possibilitats que presenta el mercat per realitzar aquest control. A més a més, també s’ha realitzat un disseny i la creació d’un prototip el qual s’instal·laria en l’interior de la nau amb la finalitat de recaptar totes les dades ambientals necessàries. Aquest prototip està comunicat amb un servidor el qual emmagatzema les dades obtingudes i les processa per tal de fer-les arribar als clients. El mètode emprat per realitzar la tasca de visualització de les dades consisteix en una pàgina web on es poden consultar diverses gràfiques i taules. Apart de la pàgina web també hi ha implementat un sistema d’alarma que avisa de manera instantània al client si alguns dels paràmetres que es controlen assoleixen un valor preocupant pel benestar de l’animal. A més a més de tot això aquest projecte també explica la programació que s’ha dut a terme per cadascun dels components per a poder assolir els objectius marcats.

Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2016, Director: Lluís Garrido Ostermann, Our world is becoming more interconnected with everything. As time passes, more electronic components emit information to the Internet. Mobile devices, intelligent wearables, beacons, any kind of machine with enough electricity to connect to the Internet may send information of its surroundings. At Beabloo, we strive to achieve a higher understanding of the environment these devices describe us by listening them. As the amount of these devices increases, harnessing such high amount of information becomes difficult, and we believe it is necessary to study new ways that allow us scale up horizontally, and allows us to bring such information as fast as possible to our data processing pipelines, which then are being used to calculate Key Performance Indicators that will be used for decision making at higher level. This paper describes a design of a data ingression architecture, keeping in mind previous design flaws that slowed down ingestion capacity. Its objective is, by keeping a small and minimalistic architecture design, to allow attaining an almost real-time data processing performance. The project uses Apache Kafka, and Apache Storm to achieve a new real-time execution environment, and to be able to evaluate its performance, it uses Prometheus to store performance metrics and Grafana to visualize them with ease.

[spa] Gaia es la misión espacial astrométrica más ambiciosa de la Agencia Espacial Europea (ESA). El satélite fue lanzado el 19 de Diciembre de 2013 y su objetivo principal es la determinación, con una resolución y precisión sin precedentes, de las posiciones, distancias y velocidades de más de mil millones de estrellas de nuestra galaxia. Esta Tesis se centra en el desarrollo del sistema de procesado IDU, "Intermediate Data Updating". IDU es una de las etapas de calibración instrumental y reducción de datos astrométricos más exigente del sistema de procesado del satélite Gaia. Sin este sistema, Gaia no podría alcanzar el nivel de precisión que se quiere obtener y su presencia es fundamental para lograr la convergencia óptima del sistema iterativo de procesado de datos en el que se basa la reducción de datos de Gaia. El procesado de los datos de Gaia es un gran reto tecnológico. En particular, el gran volumen de datos a procesar y el elevado número de procesos involucrados ha implicado el diseño de un sistema de distribución y procesado de datos muy complejo. Este procesado se basa en un sistema iterativo entre varios procesos en el que se añaden de manera continuada los nuevos datos recibidos del satélite. De entre estos procesos, esta tesis se centra en el diseño e implementación de IDU, donde se vuelven a procesar todos los datos brutos usando las calibraciones más recientes obtenidas del resto de procesos. El diseño e implementación de IDU ha supuesto una gran variedad de retos; incluyendo los problemas puramente científicos pero también las dificultades técnicas que aparecen en el procesado del gran volumen de datos de Gaia y la gestión de todas las tareas de desarrollo, test y coordinación de los equipos que contribuyen a este sistema. IDU se ejecuta en el supercomputador Marenostrum, gestionado por el "Barcelona Supercomputing Center" (BSC). Finalmente, esta tesis incluye los resultados de la primera ejecución operacional de IDU la cual ha servido para demostrar que el sistema desarrollado esta listo para afrontar el exigente reto de procesar los datos reales de Gaia durante los próximos años de misión., [eng] Gaia is an extremely ambitious astrometric space mission adopted within the scientific programme of the European Space Agency (ESA) in October 2000. It aims to measure with very high accuracy the positions, motions and parallaxes of a large number of stars and galactic objects, including also for almost all the objects information about their brightness, colour, radial velocity, orbits and astrophysical parameters. Gaia requires a demanding data processing system on both data volume and processing power. The treatment of the Gaia data has been designed as an iterative process between several systems each one solving different aspects of the data reduction system. In this thesis we have addressed the design and implementation of the Intermediate Data Updating (IDU) system. IDU is the instrument calibration and astrometric data processing system more demanding in data volume and processing power of the data processing system of the Gaia satellite data. Without this system, Gaia would not be able to provide the envisaged accuracies and its presence is fundamental to get the optimum convergence of the iterative process on which all the data processing of the spacecraft is based. The design and implementation of an efficient IDU system is not a simple task and a good knowledge of the Gaia mission is fundamental. The design and implementation of IDU is not only referring to the actual design and coding of the system but also to the management and scheduling of all the related development tasks, system tests and in addition the coordination of the teams contributing to this system. The developed system is very flexible and modular so it can be easily adapted and extended to cope with the changes on the operational processing requirements. In addition, the design and implementation of IDU presents a variety of interesting challenges; covering not only the purely scientific problems that appear in any data reduction but also the technical issues for the processing of the huge amount of data that Gaia is providing. The design has also been driven by the characteristics and restrictions of the execution environment and resources -- Marenostrum supercomputer hosted by the Barcelona Supercomputing Center (BSC) (Spain). Furthermore, we have developed several tools to make the handling of the data easier; including tailored data access routines, efficient data formats and an autonomous application in charge of handling and checking the correctness of all the input data entering and produced by IDU. Finally, we have been able to test and demonstrate how all the work done in the design and implementation of IDU is more than capable of dealing with the real Gaia data processing. We have basically executed two of the IDU tasks over the first ten months of routine operational Gaia data. This execution has been the very first cyclic data processing level run over real data so far. Executing IDU at Marenostrum over that amount of data for the first time has been a challenging task and from the results obtained we are confident that the system, we have designed and that constitutes the bulk of this thesis, is ready to cope with the Gaia data according to the requirements sets. Furthermore, the presented design provides a solid IDU system foundation for the challenging task of processing the Gaia data during the forthcoming years.

Treballs Finals de Grau de Física, Facultat de Física, Universitat de Barcelona, Any: 2015, Tutor: Jordi Torra, The main goal is to develop efficient tools to help validating the data received daily from the Gaia satellite. In order to do so, we study the main data processes and their outputs and build three different tools called ExtendedStats, FlIdtValidation and AutoAlarms.

Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2015, Director: Santi Seguí Mesquida, Every second, massive amount of data are generated around us. In an era where data is one of the most valuable resources that humans can own, the extraction of knowledge from this data has improved its relevance. Data science is science that studies the extraction of knowledge from large volumnes of data. Data visualization is one of the subjects included in data science. It’s defined as a toolkit with the goal of present this knowledge in a graphical format: a chart. Along development of this project, we’ll use a web environment to implement some of them. Therefore we’ll use some datasets whose focus is a particular topic, labor inequality, to ask ourselves some questions we’ll try to solve using upcoming knowledge and technologies which will allow us to display it.

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2015, Director: Daniel Pérez Palau, This work makes a study of eclipses, in particular solar eclipses, and how to obtain various data of an eclipse. Before it does a short introduction to astronomy, particularly the study of the positions of the stars and their movements. This allows us to better understand the movements of the Sun, the Earth and the Moon, the three objects involved in the eclipses studied. Finally we end with a chapter where we explain the implementation of what we have studied using events generated by JPL (Jet Propulsion Laboratory). Obtaining various data eclipses. Taking advantage of programming we are able to determine data with better accuracy.

Online surveys are increasingly used due to their speed and efficiency. The aim of this study was to analyze factors that may have contributed to the quality and speed of response of an online survey on influenza vaccination in primary healthcare workers. A multicenter study including family physicians, nurses and pediatricians from primary healthcare teams from seven Spanish Autonomous Communities was designed. The centers were selected by simple random sampling. The survey remained active and accessible for 56 days and four reminders were sent. The odds ratios (OR) and their 95% confidence intervals (CI) were calculated to assess the association of sociodemographic variables and responding to the survey before the second reminder. Complete, validated information was obtained from 1965 primary healthcare workers. The total response rate was 36.2%. More nurses (46.3%) responded before the second reminder and more family physicians (52.8%) after the second reminder. The adjusted OR shows that family physicians responded later (AOR 1.46, 95% CI 1.2-1.8) than nurses. The responses obtained in the first 24 h after the initial sending and the reminders accounted for 41.7% of the completed surveys, indicating the importance of reminders. This study was supported by the Ministry of Economy and Competitiveness, Institute of Health Carlos III (Grant GR09/0030), and the Catalan Agency for the Management of Grants for University Research (AGAUR Grant number 2014/SGR 1403).

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2015, Director: Jaime Nebrera i Josep Vives i Santa Eulàlia, The large amount of data stored in the last two decades makes us to create new algorithms that can treat this information and elicit the desired statistical analysis. In this paper we study the algorithms that find the most common elements k, top-k in a dataset or in a Data Stream. The algorithms must make a minimum memory usage, so the results will be an estimate, trying to minimize possible error. At the end of this work will be expose own implementation.

Podeu consultar el programari a: http://hdl.handle.net/2445/53403, Motivation: The comparative analysis of gene gain and loss rates is critical for understanding the role of natural selection and adaptation in shaping gene family sizes. Studying complete genome data from closely related species allows accurate estimation of gene family turnover rates. Current methods and software tools, however, are not well designed for dealing with certain kinds of functional elements, such as microRNAs or transcription factor binding sites. Results: Here, we describe BadiRate, a new software tool to estimate family turnover rates, as well as the number of elements in internal phylogenetic nodes, by likelihood-based methods and parsimony. It implements two stochastic population models, which provide the appropriate statistical framework for testing hypothesis, such as lineage-specific gene family expansions or contractions. We have assessed the accuracy of BadiRate by computer simulations, and have also illustrated its functionality by analyzing a representative empirical dataset.