Search

The single-index model is a non-parametric model that is popular in statistics for the estimation of risk assessment, thanks to its flexibility. Many studies have approached the problem of finding the best estimators for the model parameters and smooth function. However, most of the real-world data is asymmetric. In this thesis, we consider the application of a Box-Cox transformation on the response variable to improve the performances of the single-index model. Indeed, the Box-Cox transformation addresses the skewness of the response and transform it into a normally-distributed variables. We consider also the log transformation of the response variable. We show through a simulation study that the parameters of the models fitted with the transformed response have more similarity with the true parameters, which shows that the model is more reliable. Finally, we perform a study on a real-world data set from the car insurance industry, where we build a single-index model to predict the expected cost per claim based on a set of variables, that concern the demographic characteristic and the driving style of the customer. We assess the performance of the model fitted on the original data as well the models that use the log and the Box-Cox transformation, that obtain better results than the original one.

Les entitats financeres de crèdit necessiten obtenir previsions precises de les futures taxes d'impagament, ja que són clau per a la determinació d'estratègies adequades de gestió del risc. Unes bones previsions (és a dir, baixes taxes d'impagament) permeten a les empreses financeres ampliar la seva cartera. Per contra, la identificació d'escenaris a la baixa (és a dir, altes taxes d'impagament) condueix a una restricció de les polítiques de concessió i a augments dels fons de capital per cobrir possibles pèrdues. Atesa la relació estreta entre els impagaments en una societat i l'estat de l'economia, s'ha demostrat que les variables macroeconòmiques són forts predictors de la taxa d'impagament. Com a conseqüència de l'estructura temporal de les dades i la disponibilitat de diverses variables, s'han considerat tècniques de sèries temporals multivariants. Aquest projecte es centra en un d'aquests models: el model ARDL i l'UECM. S'han seguit dos enfocaments: el primer intenta construir un model únic, mentre que el segon construeix diversos models bivariants simples i afegeix les seves prediccions. L'últim d'aquests enfocaments ha demostrat ser millor, fins i tot superant el model de referència univariant en alguns casos. Las entidades financieras de crédito necesitan obtener previsiones precisas de sus futuras tasas de impago, ya que éstas son clave para la determinación de estrategias adecuadas de gestión del riesgo. Unas buenas previsiones (es decir, bajas tasas de impago) permiten a las empresas financieras ampliar su cartera. Por el contrario, la identificación de escenarios bajistas (es decir, altas tasas de impago) conduce a una restricción de las políticas de concesión y a aumentos de los fondos de capital para cubrir posibles pérdidas. Dada la estrecha relación entre los impagos en una sociedad y el estado de la economía, se ha demostrado que las variables macroeconómicas son fuertes predictores de la tasa de impago. Como consecuencia de la estructura temporal de los datos y la disponibilidad de distintas variables, se han considerado técnicas de series temporales multivariantes. Este proyecto se centra en uno de estos modelos: el modelo ARDL y el UECM. Se han seguido dos enfoques: el primero trata de construir un modelo único, mientras que el segundo construye varios modelos bivariantes simples y agrega sus predicciones. El último de estos enfoques ha demostrado ser mejor, superando incluso al modelo de referencia univariante en algunos casos. Credit financial entities need to obtain accurate forecasts of their future default rates as these are key for the determination of adequate risk management strategies. Good previsions (i.e. low default rates) allow financial companies to expand their portfolio. In contrast, the identification of downside/stream scenarios (i.e. high default rates) leads to a restriction of concession policies and increases in capital funds to cover potential losses. Given the close relationship between defaults in a society and the state of the economy, macroeconomic variables have been demonstrated to be strong predictors of the default rate. Given the time structure of the data and the availability of different variables, multivariate time series techniques have been considered. This project focuses on one of these models: the ARDL model and the UECM. Two approaches have been followed: the first approach tries to construct a single model while the second approach builds various simple bivariate models and aggregates their predictions. The latter of these approaches has been proven to be better, even outperforming the univariate benchmark model in some cases.

The mental health landscape worldwide has been profoundly affected by the COVID-19 pandemic, with an escalating rise in stress, anxiety, and depression. As a result, the urgency to create and apply efficacious methods for averting and addressing these mental health issues has never been greater. In this context COVICAT is a research project that is being carried out in Catalonia, Spain. The project is led by the Germans Trias i Pujol Research Institute (IGTP) and the Barcelona Institute for Global Health (ISGlobal). The goal of COVICAT is to evaluate the impact of the COVID-19 pandemic on the Catalan population. This master's thesis aims to leverage machine learning methodologies for psychological risk assessment using the COVICAT dataset (N = 9,398), contributing to healthcare practices by facilitating the initial classification of patients requiring psychological intervention. The study focuses on predicting four outcome variables reflecting patients' psychological wellbeing, one of which encapsulates a patient's self-perceived mental state. Initial exploratory data analysis and data engineering were performed to prepare the dataset for machine learning applications, including bagging, boosting, and stacking models. Each model underwent an analogous pipeline with minor variations, dictated by the distinct nature of the respective model. Post-training, these models were evaluated on unseen data, generating several performance metrics. The F1-score, designated as the critical criterion by the subject matter experts, was employed to gauge the models' efficacy. Based on these evaluations, the optimal predictive model was selected for each target. The resulting models delivered performance surpassing the adequate standard established initially by the experts (F1 = 0.6). These models aim to augment, not supplant, professional healthcare diagnoses by enabling efficient patient triage for psychological assistance.

Aquest estudi consisteix en determinar els perfils de jugadors de futbol que van disputar la Champions League entre el 2009/2010 i el 2021/2022 per tal de poder entendre les característiques i aquells factors que podrien determinar el rendiment individual i influir en l’èxit col·lectiu. En els últims anys, l'ús de les dades i l'anàlisi estadístic s'ha convertit en una part integral del món de l'esport. Les empreses i els equips han utilitzat les dades per comprendre millor els patrons i les tendències que afecten el rendiment dels esportistes i dels equips. Això ha donat lloc a l'aparició dels sports analytics, una disciplina que s'utilitza per analitzar dades esportives i obtenir informació útil per als equips, els entrenadors i jugadors. La tecnologia ha estat un dels factors clau en l'evolució dels 'sports analytics i la soccer analytics. Les dades esportives es recopilen ara de manera més precisa i detallada que mai abans, gràcies a les tecnologies com les càmeres de seguiment, els sensors i les xarxes de sensors. Això ha permès als equips obtenir dades més precises sobre el rendiment dels jugadors i sobre les condicions del joc. La creació d'una base de dades des de zero, la búsqueda de dades fiables i recolzades per la comunitat, l’estructuració de la base de dades i l’elecció dels factors claus de rendiment són, molt importants per poder obtenir resultats precisos i que s'adeqüin als objectius plantejats són els pilars fonamentals d’aquest estudi. S’han pogut determinar diversos perfils de jugadors en funció de les característiques seleccionades, com són les defensives, les ofensives, i les de construcció de joc. S’observen clares diferències entre els perfils de jugadors segons les característiques defensives proposades, mentres que les diferències entre els jugadors que formen part dels clústers de construcció de joc no són tan significatives. De la mateixa manera, els resultats obtinguts es poden interpretar de diverses maneres en funció de l’objectiu que es vulgui estudiar, i tindran un enfocament o un altre segons el punt de vista de l'anàlisi realitzat.

This paper provides a simulation and optimisation-based system to combine public transport (PT) with ride-pooling services (RP). According to the International Transport Forum (ITF), the RP could be established as a feeder of PT and included as the first or last leg of the journey with the option of transferring to/from PT in between. The system contains a dispatching core that uses an optimisation model with heuristic parameters to quickly analyse the potential permutations for each request. This topic is frequently based on simplistic modelling in the literature, and it has not been extensively tested in major urban regions. The whole metropolitan region of Barcelona is employed in this study, with a large realistic simulation model encompassing a 20 × 15 km area with a PT network of about 3000 stations and 300 route lines and nearly 114,000 traffic links. This enables for a more accurate evaluation of system performance and trip quality computation. This research was funded by an Industrial PhD by the Generalitat de Catalunya and the company PTV-AG under Grant DI-071 2019; and a research project of the Spanish R+D Programs under Grant PID2020-112967GB-C31.

[eng] Personal food security means that they have physical and economic access to sufficient, safe, and quality food. On the other hand, there are three main causes of food insecurity: 1. the high vulnerability of agriculture to climate change; 2. local, national, or international conflicts; and 3. economic inequality. The United Nations (UN) Sustainable Development Goals (SDGs), which are objectives that should be attained by 2030, are targeted at both developing and developed nations. The second SDG, "Zero Hunger," aims to double small-scale food producers' productivity and earnings while promoting resilient agricultural methods and ensuring sustainable food production. Agriculture is heavily reliant on factors related to climate change conditions such as abiotic stress, which includes soil nutrient deficiencies, accelerating temperature rises, drought, and rising CO2 concentrations; and biotic stress, which includes invasive pests, disease outbreaks, and decreased crop output. Remote sensing (RS) technologies can provide several non-destructive methods for identifying and quantifying various types of stress. For the application of RS, it is relevant to consider the different types of resolution: spectral, spatial, temporal, and radiometric. Also, the different observation scales are ground- based, aerial, space-based, or using orbital satellites. In this thesis, we evaluated the practical implementation of non-destructive methods using RS technologies across the four chapters. Moreover, we compared the maturity levels between different types of technologies using Technological Readiness Level (TRL) assessments. In the first chapter, our objective was to estimate the grain yield of the maize under low nitrogen using Vegetation Indices (VIs) from RGB (Red, Green, Blue composite color images) sensors at the ground and aerial levels in Sub-Saharan Africa (SSA). We developed an RS system in the second chapter to monitor an early warning fall armyworm (FAW) across SSA. In the third one, using leaf spectral reflectance and advanced regression models, we estimated the Vc,max, and Jmax of soybean and peanut. And in the last chapter, we developed a user-friendly mobile app for the Middle East and North Africa (MENA) countries for plant disorders detection on tomatoes, cucumbers, peppers, and quinoa, covering everything from data collection to deep learning model creation, to web and mobile app launch. Different RS technologies were used in different countries at different scales and with different types of sensors. Nevertheless, it was very relevant to consider the objectives of each study because they determined the type of spectral, spatial, and temporal resolution and the scale of observation. Regarding the TRLs across the four chapters, they suggest that the level of technology readiness depends on the goals, the time to develop the project, the amount of data collection required, and the robustness of the validation., [spa] La seguridad alimentaria personal significa que tienen acceso físico y económico a alimentos suficientes, seguros y de calidad. Por otro lado, existen tres causas principales de la inseguridad alimentaria: 1. la alta vulnerabilidad de la agricultura al cambio climático; 2. conflictos locales, nacionales o internacionales; y 3. desigualdad económica. Los Objetivos de Desarrollo Sostenible (ODS) de las Naciones Unidas (ONU), que son objetivos que deben alcanzarse para 2030, están dirigidos tanto a las naciones en desarrollo como a las desarrolladas. El segundo ODS, "Hambre cero", tiene como objetivo duplicar la productividad y los ingresos de los productores de alimentos a pequeña escala, al tiempo que promueve métodos agrícolas resilientes y garantiza la producción sostenible de alimentos. La agricultura depende en gran medida de factores relacionados con las condiciones del cambio climático, como el estrés abiótico, que incluye deficiencias de nutrientes del suelo, aumentos acelerados de temperatura, sequía y aumento de las concentraciones de CO2; y estrés biótico, que incluye plagas invasoras, brotes de enfermedades y disminución de la producción de cultivos. Las tecnologías de teledetección remota (RS) pueden proporcionar varios métodos no destructivos para identificar y cuantificar varios tipos de estrés. Para la aplicación de RS, es relevante considerar los diferentes tipos de resolución: espectral, espacial, temporal y radiométrica. Además, las diferentes escalas de observación son terrestres, aéreas, espaciales o utilizando satélites orbitales. En esta tesis, evaluamos la implementación práctica de métodos no destructivos utilizando tecnologías RS en los cuatro capítulos. Además, comparamos los niveles de madurez entre diferentes tipos de tecnologías utilizando evaluaciones de nivel de preparación tecnológica (TRL). En el primer capítulo, nuestro objetivo era estimar el rendimiento de grano del maíz bajo un nivel bajo de nitrógeno usando índices de vegetación (VI) de sensores RGB (imágenes compuestas de color rojo, verde y azul) a nivel del suelo y del aire en el África subsahariana (SSA). Desarrollamos un sistema RS en el segundo capítulo para monitorear un gusano cogollero de advertencia temprana (FAW) en SSA. En el tercero, usando reflectancia espectral de hoja y modelos de regresión avanzada, estimamos la Vc,max y Jmax de soja y maní. Y en el último capítulo, desarrollamos una aplicación móvil fácil de usar para los países de Medio Oriente y África del Norte (MENA) para la detección de trastornos de las plantas en tomates, pepinos, pimientos y quinua, que cubre todo, desde la recopilación de datos hasta la creación de modelos de aprendizaje profundo, al lanzamiento de aplicaciones web y móviles. Se utilizaron diferentes tecnologías RS en diferentes países a diferentes escalas y con diferentes tipos de sensores. Sin embargo, fue muy relevante considerar los objetivos de cada estudio porque determinaron el tipo de resolución espectral, espacial, temporal y la escala de observación. Con respecto a los TRL en los cuatro capítulos, sugieren que el nivel de preparación tecnológica depende de los objetivos, el tiempo para desarrollar el proyecto, la cantidad de recopilación de datos requerida y la solidez de la validación.

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2023, Director: Maite López Sánchez i Jerónimo Hernández-González, [en] Collaboration is increasingly occurring in virtual online communities. In them, the forms of interaction are varied, each of the virtual communities has its own way of supporting the activities it proposes, the roles the participants take, the actions they can carry out, etc. This work formalizes the generic model that characterizes virtual communities and proposes to include a conversational agent (or chatbot) within the virtual community, with the aim of being able to understand the environment and generate interaction within it. In addition, it is proposed to use a computational method based on probabilistic graphical models and machine learning. Through the inference of data corresponding to the visible interactions of the users, the variables of the model - which also includes variables that model non-visible interactions - would take on values, which would serve to predict behaviours. In particular, the virtual community of the platform decidim.barcelona is considered, which revolves around citizen participation, that is, the active intervention of citizens in public decision-making processes. The interaction spaces are analyzed and, with the aim of encouraging interaction, the incorporation of a chatbot with the ability to consult the model is proposed.

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2023, Director: Josep Fortiana Gregori, [en] In this project we study antoencoders, a machine learning tecnique used for dimensionality reduction of databases, analizing images or generating new data. We compare them with tradicional dimensionality reduction method, the principal component analysis (PCA). Even though in some fields (specially with small databases) PCA is useful we show that autoencoders can accomplish the same tasks with better results and even accomplish new ones unattainable with PCA. We prepared programs in Python implementing several versions of autoencoders, applied frequently used databases, comparing results with those obtained with PCA, when applicable.

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2023, Director: Josep Vives i Santa Eulàlia, [en] The aim of this work is to examine spatial statistics fundamentals and to study three different types of models used for flux migration forecasting: gravity models, radiation models and deep gravity models. The attention is drawn on Gaussian random fields, stationarity, kriging, Gaussian autoregression models and Markov random fields. Four gravity models, their respective four multi-linear regression models and a radiation model are implemented, all pulling the same data, to forecast migration flows within Catalonia (Spain) in 2019 between comarques by using data from these locations.

Treballs Finals del Màster de Ciències Actuarials i Financeres, Facultat d'Economia i Empresa, Universitat de Barcelona, Curs: 2022-2023, Tutores: Eva Boj del Val and Teresa Costa Cor, The aim of this work is to deepen into a stochastic variant of the classic Chain-Ladder model to calculate claims reserves of an insurance company. Specifically, we will focus on the stochastic Generalized Multivariate Chain-Ladder model (GMCL). First, the classic deterministic method will be explained. Then, we will introduce the stochastic method and develop a practical example by using both methods to see the differences in the estimations. The function MultiChainLadder of the ChainLadder package for R (R Development Core Team, 2023) will be used

En este trabajo se ha realizado una breve introducción sobre la metodología Regresión de Proceso Gaussiano (GPR) y dos aplicaciones en el ámbito Actuarial. Por un lado, se ha realizado un ejercicio de interpolación sobre las tablas de mortalidad PASEM Unisex 2020, concluyendo que el GPR es una excelente herramienta de interpolación, y que nos permite una tarificación más ajustada en el ramo de Vida. Por otro lado, se ha integrado el GPR como medida de predicción de provisiones en los ramos de No-Vida, obteniendo unos resultados prometedores. Por último, se concluye que un GPR puede ser un instrumento útil, siempre y cuando, se realice una buena selección del Kernel y un correcto período de entrenamiento del modelo.

[eng] Chemical exposure to emerging contaminants (ECs) is a major concern nowadays. These ECs have recently become a global environmental threat, and an in-depth characterization of their occurrence and toxic impact is needed. In this context, omic sciences have arisen as powerful tools to shed some light on the biological mechanisms affected by exposure to these chemicals. Particularly, metabolomics and lipidomics can provide a snapshot of what is actually happening at the molecular level, pointing to metabolic pathways affected by the contaminants. New analytical methodologies are required to extract the sought information in more complex biological matrices (from single cells to whole organisms). Hence, a major emphasis has been put on developing multidimensional separations and multiplatform approaches to increase the metabolome coverage. However, these novel approaches bring about massive datasets, and the complexity of the data analysis augments considerably. Therefore, chemometric strategies are a perfect match to get through this bottleneck and provide useful tools to obtain the most from the data collected. In this PhD Thesis, the focus was set on developing analytical protocols, especially using two-dimensional liquid chromatography coupled to mass spectrometry (LC×LC-MS), as well as chemometric data analysis strategies applicable to environmental metabolomic studies. On the one hand, LC×LC-MS methods have been developed for both untargeted and targeted analyses. Active modulation strategies have been also successfully implemented in the multidimensional chromatographic separation of lipids. On the other hand, the Regions Of Interest (ROI) approach for compression and filtering has been validated for LC×LC-MS analyses. Regarding chemometric resolution methods (i.e., which allow obtaining quantitative and qualitative information from the sample constituents), and due to deviations from an ideal trilinear behavior presented by LC×LC datasets, the use of the Multivariate Curve Resolution Alternating Least Squares (MCR-ALS) method has been preferred. Different quantification strategies have been tested based on the Regions Of Interest Multivariate Curve Resolution (ROIMCR) approach. In addition, several multivariate statistical methods based on the analysis of variance (ANOVA) have been compared for metabolomic studies. As a result, a combination of ANOVA-simultaneous component analysis (ASCA) and partial least squares discriminant analysis (PLS-DA) has been selected for statistical analysis and variable (metabolite) selection, respectively. All in all, different metabolomic workflows have been validated for the assessment of emerging contaminants in model biosystems., [spa] Actualmente, las ciencias ómicas han aparecido como herramientas muy útiles para arrojar luz sobre los mecanismos biológicos y rutas metabólicas que se ven afectados debido a las exposiciones a contaminantes emergentes. En concreto, la metabolómica y la lipidómica proporcionan información de lo que está ocurriendo a nivel molecular. Por tanto, se requieren nuevas metodologías que sean capaces de extraer dicha información en matrices cada vez más complejas (desde una única célula a un organismo entero). Esta Tesis doctoral se centra principalmente en el desarrollo de protocolos analíticos basados en el uso de la cromatografía líquida bidimensional acoplada a espectrometría de masas (LC ×LC-MS), así como en el desarrollo de estrategias quimiométricas que permitan su uso en aplicaciones medioambientales. Por un lado, se han optimizado métodos LC×LC-MS para análisis dirigidos y no dirigidos, implementando con éxito estrategias de modulación activa (en el caso de análisis de lípidos). Por otro lado, se ha validado la estrategia de regiones de interés (ROI) para comprimir y filtrar los datos obtenidos con LC×LC-MS. Asimismo, se ha preferido el uso de métodos de resolución multivariante de curvas mediante mínimos cuadrados alternados (MCR-ALS) para la resolución cualitativa y cuantitativa de muestras complejas en el caso de datos de LC×LC, debido a las deviaciones de la trilinealidad encontradas en dichos datos. Por otra parte, se han comparado diferentes estrategias cuantitativas aplicables a datos en LC×LC-MS, todas ellas basadas en el uso del método combinado de regiones de interés y resolución multivariante de curvas (ROIMCR). También se han comparado diversos métodos estadísticos multivariante basados en el análisis de varianza (ANOVA) y su aplicabilidad en estudios metabolómicos. Finalmente, se ha elegido una combinación de análisis estadístico efectuado con ANOVA-análisis de componentes simultáneos (ASCA) y un método de clasificación, análisis discriminante mínimos cuadrados parciales (PLS-DA) para seleccionar las variables (metabolitos) más relevantes. En resumen, se han validado diferentes flujos de trabajo para el estudio metabolómico del efecto de contaminantes emergentes en organismos modelo ambientales.

[cat] En els darrers anys, l’augment en la producció de compostos orgànics sintètics i el seu constant alliberament al medi ambient, sobretot en referència als productes farmacèutics, compostos en els quals s’ha centrat aquesta tesi, ha fet que aquestes substàncies d’origen generalment antropogènic s’hagin anat acumulant als ecosistemes biològics de forma important, extensa i persistent, especialment en el medi aquàtic. Per aquest motiu, és necessari utilitzar processos de depuració ambiental per tal d’eliminar les substàncies nocives per a la salut i el medi ambient que aquestes puguin ocasionar. Tanmateix, les tecnologies de depuració emprades encara no permeten eliminar aquestes substàncies totalment de forma quantitativa. A més, els possibles efectes perjudicials d’aquestes substàncies pel medi ambient no es produeixen només pels compostos originals, sinó que en molts casos, també pels seus productes de degradació i metabòlits. Consegüentment, és important conèixer el destí i la transformació que sofreixen els contaminants orgànics un cop entren en l’ecosistema i arriben al medi aquàtic. Cal, doncs, tenir present que l’aigua és un recurs imprescindible per a la vida que requereix de la millor qualitat sanitària possible per al seu consum. La matèria orgànica dissolta (DOM), present tant en aigües superficials com en subterrànies, i que té un efecte significatiu en els processos bioquímics dels sistemes aquàtics, actua moltes vegades com indicador del nivell de qualitat i salubritat que té l’aigua. Per tant, a l’hora d’optimitzar els processos de purificació realitzats, és també de gran importància conèixer i actuar sobre la concentració i les característiques de la DOM present en ella. Per aquest motius, en la present Tesi s’han realitzat els següents estudis: D’una banda, per tal de conèixer i entendre millor el procés de fotodegradació de fàrmacs, s’ha estudiat el procés de degradació de dos d’ells. En un primer estudi, diferents mostres d’un medicament citostàtic (tamoxifè) es sotmeten a una font de radiació UV controlada en el laboratori, que simula la radiació solar ambiental (suntest), emprant dues potències d’irradiació diferents. Al mateix temps, es realitza el seguiment de la seva fotodegradació a partir de mesures espectrofotomètriques UV-Visible i de fluorescència multidimensional (matrius d’excitació-emissió, EEM). Seguidament, diferents alíquotes de mostra al llarg de la degradació són analitzades mitjançant LC-DAD, LC-MS i/o LC-FLD. Aquestes anàlisis, que generen una quantitat de dades molt elevada, es fusionen i tracten simultàniament a partir de procediments quimiomètrics avançats (especialment amb el mètode de resolució multivariant de corbes per mínims quadrats alternats (MCR-ALS)), que permeten descriure el procés i la reacció de degradació, resoldre els diferents productes de transformació (TPs), fer la seva identificació i fer l’avaluació de la cinètica associada al procés estudiat. Com a resultat d’aquest estudi, 3 TPs i 1 isòmer del tamoxifè van ser caracteritzats. Un segon estudi amb el mateix objectiu va ser realitzat sobre l’antibiòtic sulfametoxazol, present en rius i efluents de tot Europa en concentracions destacables. Addicionalment, un estudi de les propietats àcid-base d’aquesta substància va ser dut a terme per tal de veure com el pH pot afectar a l’especiació d’aquest compost. De la mateixa manera, per tal de conèixer si la fotodegradació del fàrmac es troba influenciada per aquest paràmetre, el procés de degradació va ser estudiat emprant solucions de mostra a pH diferents. Com a resultat de l’anàlisi quimiomètrica de les dades d’aquest segon estudi, un isòmer conformacional, un fotoproducte desconegut i 4 TPs van ser caracteritzats. D’altra banda, 3 formes diferents de l’antibiòtic van ser detectades en la corresponent valoració àcid-base. D’altra banda, per conèixer la qualitat de les aigües superficials dels rius, diferents mostres d’aigua superficial procedents de la conca del riu Llobregat han estat analitzades i caracteritzades mitjançant la fluorescència multidimensional (EEM). En aquest estudi, el seguiment de la qualitat de l’aigua a partir de la combinació de les mesures fluorimètriques d’EEM i de la seva modelització quimiomètrica flexible, per mitjà també del MCR-ALS, ha permès obtenir informació sobre la naturalesa de les fraccions de la DOM (origen húmic, fúlvic, proteínic, etc.) presents al riu Llobregat i de la seva distribució geogràfica al llarg de la conca en diferents campanyes de mostreig ambientals. Així doncs, atès que avui dia existeix una necessitat urgent per desenvolupar noves eines de tractament de dades multivariants, tals com les que s’han obtingut en aquesta Tesi en els diferents casos d’interès ambiental estudiats, la memòria que es presenta vol contribuir a resoldre aquest coll d'ampolla a partir de l’ús de la quimiometria, i en especial del mètode MCR-ALS, el qual pot jugar un paper molt important en la resolució de molts dels problemes associats a l’anàlisi de dades en diferents àmbits d’aplicació química i ambiental., [eng] In recent years, the increase in the production of synthetic organic compounds and their constant release into the environment, especially in reference to pharmaceuticals, compounds on which this thesis has focused, has meant that these substances of origin generally anthropogenic have been accumulating in biological ecosystems in an important, extensive, and persistent way, especially in the aquatic environment. For this reason, it is necessary to use environmental purification processes to eliminate substances that are harmful to health and the environment. However, the purification technologies used today do not yet allow these substances to be completely eliminated in a quantitative way. In addition, the potentially harmful effects of these substances on the environment are not only due to the original compounds, but in many cases also to their degradation products and metabolites. Consequently, it is important to know the fate and transformation that organic pollutants undergo once they enter the ecosystem and reach the aquatic environment. We must therefore keep in mind that water is an essential resource for life that requires the best possible sanitary quality for its consumption. Dissolved organic matter (DOM), present in both surface and groundwater, and which has a significant effect on the biochemical processes of aquatic systems, often acts as an indicator of the level of quality and health of water. Therefore, when optimizing the purification processes carried out, it is also very important to know and act on the concentration and characteristics of the DOM present in it. For this reason, the following studies have been performed in this Thesis: On the one hand, to better understand the process of photodegradation of drugs, the degradation process of two of them has been studied. In a first study, different samples of a cytostatic drug (tamoxifen) were subjected to a controlled source of UV radiation in the laboratory, which simulates solar radiation (suntest), using two different irradiation powers. At the same time, its photodegradation is monitored using UV-Visible spectrophotometric measurements and multidimensional fluorescence (excitation-emission matrices, EEM). Then, different sample aliquots along the degradation are analyzed by LC-DAD, LC-MS and / or LC-FLD. These analyzes, which generate a very high amount of data, are merged and treated simultaneously using advanced chemometric procedures (especially with multivariate curve resolution alternating least squares method (MCR-ALS)), which describe the process and the degradation reaction, solve the different transformation products (TPs), identify them and evaluate the kinetics associated with the process studied. As a result of this study, 3 TPs and 1 isomer of tamoxifen were characterized. A second study with the same aim was carried out on the antibiotic sulfamethoxazole, present in rivers and effluents from all over Europe in remarkable concentrations. Additionally, a study of the acid-base properties of this substance was conducted to see how pH can affect the speciation of this compound. Similarly, to know whether the photodegradation of the drug is influenced by this parameter, the degradation process was studied using sample solutions at different pH. As a result of the chemometric analysis of the data from this second study, a conformational isomer, an unknown photoproduct, and 4 TPs were characterized. On the other hand, 3 different forms of the antibiotic were detected in the corresponding acid-base titration. On the other hand, to know the quality of the surface water of the rivers, different samples of surface water from the Llobregat river basin have been analyzed and characterized by multidimensional fluorescence (EEM). In this study, the monitoring of water quality from the combination of the fluorometric measurements of EEM and its flexible chemometric modeling, also through the MCR-ALS, has made it possible to obtain information on the nature of the fractions of the DOM (humic, fulvic, protein origin, etc.) present in the Llobregat river and its geographical distribution along the basin in different environmental sampling campaigns. Therefore, given that there is an urgent need today to develop new multivariate data processing tools, such as those obtained in this Thesis in the different cases of environmental interest studied, the presented report aims to contribute to solve this bottleneck from the use of chemometrics, and especially the MCR-ALS method, which can play a very important role in solving many of the problems associated with data analysis in different areas of chemical and environmental application.

Treballs Finals de Grau d'Enginyeria Informàtica, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2022, Director: Sergio Escalera Guerrero i Carles Casacuberta, [en] Memes evolve and mutate through their diffusion in social media. They have the potential to propagate ideas and, by extension, products. Many studies have focused on memes, but none so far, to our knowledge, on the users who post them, their relationships, and the reach of their influence. In this project, we define a meme influence graph together with suitable metrics to visualize and quantify influence between users who post memes. Then, we describe a process to implement our definitions using a new approach to meme detection based on text- to-image area ratio and contrast. After applying our method to a set of users of the Instagram platform, we conclude that our metrics add information to already existing user characteristics and that our methodology can also be used to study the popularity of memes types among the users.

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2022, Director: Sergi Baena i Miret, [en] In this project we will show and discuss the classification algorithms, specifically, for the breast cancer diagnosis. From a theoretical point of view, we will study and prove the basic results of multivariate analysis, such as: dimension theorem, properties of multivariate distributions and the necessary results of Principal Components Analysis (PCA) with their respectively proofs. Then, from a more practical point of view, we will present the observed data, understanding their meaning, studying their properties and the subsequent application of a PCA. Finally, using R programming language, we will apply the data to the classification algorithms Naive Bayes and Support Vector Machine, showing the results that they provide. As well as we will see a brief explanation of the K-NN algorithm.

El objetivo de este trabajo es estudiar la asociación entre las medidas antropométricas y la actividad física, y el riesgo de desarrollar un cáncer de cuello uterino o uno de sus precursores. A partir de la base de datos del estudio EPIC (European Prospective Investigation into Cancer and Nutrition) se realizarán varios tipos de análisis: modelos de riesgo proporcional de Cox para estimar hazard ratio y modelos de regresión logística condicional para estimar odds ratios. Además, se validarán ambos tipos de modelos, en concreto, el de regresión logística condicional mediante un ajuste de bondad específico para datos emparejados.

En este documento se intenta proporcionar aquellos elementos básicos, tanto teóricos como prácticos para poder establecer y comprobar estructuras de datos a partir de escalas de respuesta. De forma específica se explicitan tanto los aspectos fundamentales de la fiabilidad y la validez y cómo estos se establecen con los programas estadísticos de análisis factorial exploratorio y confirmatorio. El software de soporte utilizado son los programas SPSS, el FACTOR.12 y AMOS.

Even though the logratio methodology provides a range of both generic, mostly exploratory, and purpose-built coordinate representations of compositional data, simple pairwise logratios are preferred by many for multivariate analysis in the geochemical practice, principally because of their simpler interpretation. However, the logratio coordinate systems that incorporate them are predominantly oblique, resulting in both conceptual and practical problems. We propose a new approach, called backwards pivot coordinates, where each pairwise logratio is linked to one orthogonal coordinate system, and these systems are then used together to produce a concise output. In this work, principal component analysis (PCA) and regression with compositional explanatory variables are used as primary methods to demonstrate the methodological and interpretative advantages of the proposal. In the applied part of this study, sediment compositions from the Jizera River, Czech Republic, were analysed using these techniques through backwards pivot coordinates. This allowed to discuss grain size control of the element composition of sediments and clearly distinguish anthropogenically contaminated and uncontaminated strata in sediment depth profiles

This study presents a comprehensive bibliometric analysis of the paper published by John Aitchison in the Journal of the Royal Statistical Society. Series B (Methodological) in 1982. Having recently reached the milestone of 35 years since its publication, this pioneering paper was the first to illustrate the use of the methodology "Compositional Data Analysis" or "CoDA". By October 2019, this paper had received over 780 citations, making it the most widely cited and influential article among those using said methodology. The bibliometric approach used in this study encompasses a wide range of techniques, including a specific analysis of the main authors and institutions to have cited Aitchison' paper. The VOSviewer software was also used for the purpose of developing network maps for said publication. Specifically, the techniques used were co-citations and bibliographic coupling. The results clearly show the significant impact the paper has had on scientific research, having been cited by authors and institutions that publish all around the world The article was supported by the Spanish Ministry of Science, Innovation and Universities (grant RTI2018-095518-B-C21)

The companies' investment and financing policies are dynamically interrelated and there is no general consensus about the direction of this relationship. There are theoretical arguments and empirical evidence supporting both possible directions, which makes panel vector autoregressive models an appropriate tool. However, the financial ratios normally used to assess this relationship empirically tend to be asymmetric, and to have extreme outliers and non-linear relationships. The aim of this article is to propose a methodological approach to address these issues by complementing panel vector autoregressive models with compositional data analysis. The usefulness of the proposed methodology is illustrated with real data of Spanish retail companies, while a reanalysis with standard financial ratios is inconclusive This work was supported by the Spanish Ministry of Science, Innovation and Universities/FEDER [grant number RTI2018–095518–B–C21]; the Spanish Ministry of Health [grant number CIBERCB06/02/1002]; and the Government of Catalonia [grant number 2017SGR656]

While most countries have networks of stations for monitoring pollutant concentrations, they do not cover the whole territory continuously. Therefore, to be able to carry out a spatial and temporal study, the predictions for air pollution from unmeasured sites and time periods need to be used. The objective of this study is to predict the air pollutant concentrations of PM10, O3, NO2, SO2 and CO in sites throughout Catalonia (Spain) and time periods without a monitoring station. Compositional data (CoDa) studies the relative importance of pollutants. A novel feature in this article is combining CoDa with an indicator of total pollution. Predictions are then made using a combination of spatio-temporal models and the Bayesian Laplace Integrated Approach (INLA) inference method. The most relevant results obtained indicate that the log-ratio between NO2 and O3 has the highest variance and the best predictive accuracy in time and space. Total pollution levels are second in variance but have low spatial predictive accuracy. Third in variance is the low temporal accuracy found in the log-ratio between SO2 and the remaining pollutants. Globally, the combination of CoDa and the INLA method is suitable for making effective spatio-temporal predictions of air pollutants This work was partially financed by the SUPERA COVID19 Fund, from SAUN: Santander Universidades, CRUE and CSIC; by the COVID-19 Competitive Grant Program from Pfizer Global Medical Grants; by the Spanish Ministry of Science, Innovation and Universities/FEDER (grant number RTI2018–095518–B–C21); and the Government of Catalonia (grant number 2017SGR656) Open Access funding provided thanks to the CRUE-CSIC agreement with Elsevier

Treballs Finals del Màster de Ciències Actuarials i Financeres, Facultat d'Economia i Empresa, Universitat de Barcelona, Curs: 2021-2022, Tutor: Samer Ajour El Zein, El presente trabajo consiste en un análisis del impacto de las noticias de prensa sobre la volatilidad de Bitcoin. Se realiza un análisis gráfico tomando una base de datos de noticias importada de un importante periódico económico nacional, y se observa el impacto de ciertas palabras clave sobre la serie de precios de Bitcoin. El análisis se complementa con un análisis estadístico basado en los modelos GARCH, y concretamente utilizando el modelo multivariante DCC. Este modelo muestra que las variables de estudio se correlacionan con la variable principal, concluyendo que sí es posible estimar el impacto que pueden tener las noticias sobre esta cryptomoneda.

Objective Direct cortical stimulation (DCS) is standard for intracranial presurgical evaluation in drug-resistant epilepsy (DRE). Few studies have reported levels of concordance between spontaneous seizure generators and triggered seizures during DCS. The present work reports validity measures of DCS for detecting the seizure onset zone (SOZ) during stereoelectroencephalography (SEEG). Methods We evaluated all patients who underwent SEEG evaluation at our epilepsy center between 2013 and 2019. Data were analyzed using contingency tables. Validity measures of the diagnostic test were computed for all patients evaluated with DCS and for seizure free patients. Results Fifty-eight consecutive patients were evaluated through DCS. One hundred seventy-three clinical seizures were elicited with DCS. Electroclinical identical to spontaneous seizures were considered true positive (TP) seizures. They showed a high specificity (96.9%) for detecting the SOZ in patients that remained seizure free one year after treatment. Sensitivity was low (23.0%), and a high percentage of false-negative stimulations was documented in the SOZ. The accuracy was 87.9%. Conclusions DCS is a technique with high specificity but a low sensitivity for the localization of the SOZ. The DCS validity measures need to be known when considered for surgical decisions. The interpretation of DCS-triggered seizures and the differentiation of true-positive vs false-positive seizures should be carefully evaluated. Significance DCS seizure triggering is highly specific for SOZ localization. This work received support from the project “Clúster Emergent del Cervell Humà” (CECH) and the European Regional Development Fund under the framework of the ERFD Operative Programme for Catalonia.

We propose a new approach for the visual inspection of interactions between several variables related to the wind and solar energy sector, and a set of socioeconomic variables and natural factors that may be affecting the sector. Focusing on fifteen European countries in the period from 2007 to 2019, we use Categorical Principal Component Analysis to reduce our data into two factors, the first relating to energy consumption, greenhouse gas emissions, per capita income, and solar energy, and the second capturing climactic and sociopolitical factors. The dimensionality reduction also displays a decoupling between natural factors and variable renewable energy sources (VRES) development, particularly in the case of solar energy, and instead shows a more influential relationship with economic factors. We additionally project all countries into a perceptual map and observe three clusters that roughly correspond to the main European regions (Southern and Eastern Europe, Northern Europe and Western Europe). Finally, we plot the average level and growth level of both the wind and solar energy share for each nation and observe a negative relationship in wind share and a slightly positive relationship in solar share. Our results show that, especially for wind energy, countries with higher levels of overall renewable energy development are more likely to show more intense VRES development than countries who already have high existing levels of the technology in their renewable energy mix. Solar energy investment on the other hand is more likely to be dominated by countries with pre-existing high levels of solar in the renewables mix. Our results emphasize the importance of individual nations attitude towards renewables as a whole as playing a key role in VRES development, as much as the natural resource availability of these energies.

Treballs Finals de Grau en Estadística UB-UPC, Facultat d'Economia i Empresa (UB) i Facultat de Matemàtiques i Estadística (UPC), Curs: 2021-2022, Tutor: Lesly María Acosta Argueta, En la actualidad, la temática de la salud mental está adquiriendo mayor relevancia, pero aún se tienen pocos registros y estudios al respecto. A través del presente trabajo, que analiza estadísticamente la dinámica natural del número de suicidios en España, se quiere contribuir a dar un enfoque científico a la temática y también a que la ideación suicida tenga la importancia y visibilidad que merece. El objetivo estadístico de este estudio es analizar, con fines predictivos, la evolución de los datos de suicidios mensuales en España a lo largo del tiempo entre 1990 y 2019. Para el análisis estadístico, se han obtenido registros de suicidios del Instituto Nacional de Estadística (INE) y se ha recurrido al análisis de series temporales. Concretamente, se ha aplicado la metodología ARIMA (autoregressive integrated moving average) clásica de Box-Jenkins y su extensión conocida como ARIMAX. El procedimiento de construcción de un modelo ARIMAX, ha permitido seleccionar al final un modelo con buena calidad de ajuste y capacidad predictiva. Además, los resultados obtenidos permiten predecir con cierta confianza el comportamiento que podría seguir el número de personas que deciden quitarse la vida si no se implantan medidas de prevención efectivas.

Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2022, Director: Josep Fortiana Gregori, [en] In a statistical context, we can define a mixture model as a probabilistic model used to represent the presence of subpopulations within the same population. However, we can also say that a mixture model corresponds to a distribution (formed by a convex linear combination of other distributions) that represents the probability distribution of an observation in a population. Mixture models are used to create statistical inferen- ces, approximations, and predictions about the properties of subpopulations based on observations made about the population studied, without the need to identify the corresponding subpopulation of each observation. In this project, we will study a particular case of mixture models: the Gaussian mixture model (mixture of multivariate Gaussian distributions). The EM algorithm is a method that allows us to estimate the parameters of a statistical model when the data is incomplete or when the model contains unknown variables. In the case of mixture models, the unknown variables are those that tell us which component generated each observation in the sample. In this project, we will study the EM algorithm from different points of view, and we will use it to estimate the parameters of the Gaussian mixture model.