Pompeu Fabra University

UPF Digital Repository
Not a member yet
    43569 research outputs found

    Deep learning-based photovoltaic energy forecasting with ground-based sky imagery and atmospheric data

    No full text
    Treball fi de màster de: Master in Intelligent Interactive SystemsSupervisor: Antonio AgudoDue to the rise in the use of renewable energies as an alternative to traditional energies, and especially solar energy, there is increasing interest in studying how to address photovoltaic forecasting in the face of the challenge of variability in photovoltaic energy production, using different techniques and methodologies. This work develops a hybrid methodology for short and long-term forecasting based on two studies with the same purpose. A multimodal approach is proposed that combines images of the sky and photovoltaic energy history with meteorological data. The main objective is to improve the accuracy of ramp event prediction, increase the robustness of forecasts in cloudy conditions, and extend capabilities beyond nowcasting, in order to support more efficient operation of the power grid and better management of solar variability. Deep convolutional neural network architectures are used for nowcasting and forecasting models, incorporating individual and multiple meteorological variables, as well as solar position. The results demonstrate that the inclusion of meteorological data, particularly the surface long-wave (thermal), radiation downwards (strd), and the combination of wind and solar position, significantly improves predictions in both nowcasting and forecasting tasks, especially on cloudy days. This study highlights the importance of integrating diverse data sources to improve the reliability and interpretability of solar energy prediction models

    A novel Spanish dataset for financial education text simplification targeting visually impaired individuals

    No full text
    Automatic Text Simplification (ATS) is a crucial task in natural language processing, aimed at making texts more comprehensible, particularly for specific groups such as individuals with visual impairments. One of the primary challenges in developing models for ATS is the scarcity of data, especially in Spanish. This manuscript introduces a novel dataset tailored for Spanish speakers with visual impairments, consisting of 5,314 pairs of original and simplified sentences created using established simplification rules. Additionally, we evaluate the feasibility of augmenting this dataset using large language models such as Generative Pre-training Transformer (GPT)-3, TUNER, and Multilingual T5 (mT5). We compare the simplifications generated by these models with our dataset to assess their effectiveness in data augmentation. The characteristics of our dataset and the findings from these comparisons are discussed in detail. The dataset is publicly available on Hugging Face at https://huggingface.co/datasets/saul1917/FEINA.The work of Horacio Saggion was supported in part by the Maria de Maeztu Units of Excellence Program, funded by MCIN/AEI/10.13039/501100011033 under Grant CEX2021-001195-M; and in part by European Union’s Horizon Europe Research and Innovation Program through the iDEM Project under Grant 101132431

    The use of CUSUM chart method for surveillance of learning effects and quality of care in endovascular procedures

    No full text
    Introduction: Quality of care and learning effect surveillance are two mandatory responsibilities within a changing therapeutical paradigm. We aimed to evaluate the feasibility and value of CUSUM chart method in assessing performance in consecutive endovascular procedures done by vascular surgeons of a single department on aorto-iliac, femoropopliteal and renal artery occlusive disease. Material and method: Data were collected in 405 consecutive patients, scheduled for endovascular intervention of aorto-iliac (n = 131, 32.3%), femoropopliteal (n = 142, 35%) and renal artery (n = 132, 32.7%) occlusive disease during a 6-year period. Quality indicators included inability to cross the lesion, peri- and post-procedural complications and significant residual stenosis or occlusion at 1 month. CUSUM curves were generated for each territory globally and according to each quality indicator. The relevance of curve upward inflections was evaluated with Fisher's Exact Test. Results: Failure to cross the lesion occurred in 6.9% (aorto-iliac), 10.6% (femoropopliteal) and 2.3% (renal) of patients. One-hundredth twenty aorto-iliac, 127 femoropopliteal and 132 renal angioplasties were finally performed. Peri- and post-procedural complications appeared in 14.5% (aorto-iliac), 9.2% (femoropopliteal) and 2.3% (renal), while significant residual stenosis or occlusion was seen in 0.8%, 4.9% and 2.3% of patients, respectively. Aorto-iliac CUSUM curve showed two upward inflections at the beginning and the end of the period, both associated with peri- and post-procedural complications (p = 0.002 and p = 0.0013) and the latter also with failure to cross the lesion (p = 0.009). Femoro-popliteal CUSUM curve moved progressively upward during all the period, initially related to peri- and post-procedural complications (p = 0.038) and later to failure to cross the lesion (p = 0.004). Renal CUSUM curve didn't show any upward inflection during the analysed period. Conclusion: CUSUM curves are an excellent tool for measuring learning effect and quality of care within a changing paradigm, such it is the case of endovascular interventions. Curve upward inflections can be further interpreted according to the type of "failure" thus helping to evaluate their underlying causes

    La Traducción español-chino para subtitulado de elementos de suspense

    No full text
    Este estudio analiza las soluciones de los subtítulos español-chino para los ELSU (elementos de suspense) en películas del género thriller, fundamentándose en la teoría de la traducción audiovisual y estudios previos sobre el suspense. Se centra en cinco películas de thriller: Cien años de perdón (2016), Contratiempo (2017), El bar (2017), Durante la tormenta (2018), y Al final del túnel (2019). Utilizando una metodología descriptiva de la traducción audiovisual, se evidencia la expresión verbal, no verbal y paraverbal del suspense, así como las técnicas de suspense empleadas en estas películas. Posteriormente, se describen las soluciones subtituladas aplicadas para estos ELSU y técnicas de suspense mediante árboles binarios, para analizar las relaciones entre la creación del suspense en el texto de partida y su recreación mediante la traducción audiovisual en el texto de llegada. Este estudio aporta un avance en la comprensión de la subtitulación interlingüística español-chino acerca de los ELSU, dilucidando cuestiones asociadas con el suspense y desarrollando una metodología de análisis multimodal para abordar problemas en la traducción audiovisual.This thesis investigates Spanish-Chinese subtitle solutions for elements of suspense (ELSU) in thriller films, grounding its analysis in audiovisual translation theory and previous studies on suspense. Using a descriptive approach to audiovisual translation, this study delves into five thriller films: Cien años de perdón (2016), Contratiempo (2017), El bar (2017), Durante la tormenta (2018), y Al final del túnel (2019). It identifies and analyzes verbal, non-verbal, and paraverbal expressions of suspense, along with the suspense techniques employed in these films. The research further describes and outlines the subtitled solutions applied to these suspense elements and techniques, using binary trees to explore the relationships between the creation of suspense in the source text and its recreation through audiovisual translation in the target text. This study advances the understanding of Spanish-Chinese interlingual subtitling with respect to ELSUs, elucidates the associated challenges with suspense, and develops a multimodal analysis methodology to tackle issues in audiovisual translation.Aquest estudi analitza les solucions de subtitulació espanyol-xinès per als ELSU (elements de suspens) en pel·lícules del gènere thriller, basant-se en la teoria de la traducció audiovisual i estudis previs sobre el suspens. Es centra en cinc pel·lícules de thriller: Cien años de perdón (2016), Contratiempo (2017), El bar (2017), Durante la tormenta (2018) i Al final del túnel (2019). Mitjançant una metodologia descriptiva de la traducció audiovisual, es mostra l'expressió verbal, no verbal i paraverbal del suspens, així com les tècniques de suspens emprades en aquestes pel·lícules. Posteriorment, es descriuen les solucions de subtitulació aplicades a aquests ELSU i tècniques de suspens mitjançant arbres binaris, per analitzar les relacions entre la creació del suspens en el text de partida i la seva recreació mitjançant la traducció audiovisual en el text d'arribada. Aquest estudi aporta un avenç en la comprensió de la subtitulació interlingüística espanyol-xinès sobre els ELSU, ja que permet aclarir els problemes associats amb el suspens i desenvolupa una metodologia d'anàlisi multimodal per abordar problemes en la traducció audiovisual.Programa de Doctorat en Traducció i Ciències del Llenguatg

    Fuel-based lighting and under-five morbidity in semi-rural Mozambique: A cohort study

    No full text
    Background: Mozambique has one of the lowest electrification rates in the world, leaving its population dependent on polluting fuels for lighting. Limited epidemiological evidence links fuel-based lighting to child health. We examined associations between polluting lighting sources, particularly kerosene, and under-five morbidity in a semi-rural district in Mozambique. Methods: We constructed a birth cohort from demographic and hospital-based pediatric morbidity surveillance data. We included children born in Manhiça district between January 1, 2016, and December 31, 2020. The most common lighting source (polluting vs clean) used during follow-up was used as exposure. The outcome was the frequency of all-cause and respiratory-linked hospital visits in any of the seven surveilled health facilities within the district. We used zero-inflated negative binomial regression models to calculate Rate Ratios (RR) adjusted for potential confounders at the child, mother, and household level. Results: We included 17,815 under-five children (49% female) living in 13,574 households. Nearly a quarter of children (24.7%) lived in households using polluting lighting fuels. During follow-up, there were a total of 69,677 all-cause hospital visits (53% respiratory-linked). Children in households with polluting lighting had a 2% higher rate of all-cause hospital visits (RR: 1.02, 95% CI: 0.98, 1.06) and a 2% lower rate of respiratory-linked hospital visits (RR: 0.98, 95% CI: 0.94, 1.03) compared to those with clean lighting. Results were robust to sensitivity analyses testing alternative inclusion criteria based on the definition of health facility catchment area and duration of follow-up time covered by the surveillance system. Conclusion: Polluting compared to cleaner lighting sources were not associated with hospital visits among children under five. Intervention-based research is needed to better understand the health impacts of air pollution from fuel-based lighting among children living in areas with limited access to electricity

    Los resultados de la evaluación PISA del año 2022 en Cataluña: ¿un problema político o educativo?

    No full text
    Treball de Fi de Grau en Ciències Polítiques i de l'Administració. Curs 2024-2025Tutora: Mireia López ÁlvarezEste trabajo analiza cómo los resultados de la evaluación PISA 2022 en Cataluña se han transformado en un problema político, trascendiendo su dimensión educativa. A partir de un enfoque mixto, se combina el análisis cuantitativo de los datos PISA y del CSASE —que muestran un bajo rendimiento medio, alta dispersión y fuertes desigualdades según nivel socioeconómico, origen migrante y tipo de centro— con un enfoque cualitativo sobre su incorporación en la agenda política catalana. Utilizando la teoría de las “tres corrientes” de Kingdon, se explica cómo factores como la cobertura mediática, las disputas ideológicas o la presión de actores institucionales han facilitado esta incorporación. Los resultados que obtiene Cataluña en PISA 2022 nos conducen a preguntarnos porqué el sistema educativo catalán presenta mayor dispersión en las notas medias (menor homogeneidad) y la mayor diferencia entre las notas más altas y las más bajas cuestiona si es tan equitativo (en comparación con el resto de CCAA). El estudio identifica que la equidad y la homogeneidad son los principales desafíos del sistema educativo catalán y que existen márgenes de actuación dentro del marco competencial de la Generalitat. Se proponen políticas públicas basadas en la evidencia, aplicadas a centros educativos públicos y focalizadas —como programas de tutorización, financiación diferenciada (pagos por rendimiento), mayor autonomía con rendición de cuentas y estrategias de comunicación institucional— que permitirían mejorar los resultados y, paralelamente, desactivar su centralidad política. Finalmente, se sugieren líneas futuras de investigación sobre la segmentación educativa, el impacto de la inmigración y los efectos de la inmersión lingüística desagregados por nivel socioeconómico

    Epithelial cells provide immunocompetence to the early embryo for bacterial clearance

    No full text
    Early embryos are exposed to environmental perturbations that may influence their development, including bacteria. Despite lacking a proper immune system, the surface epithelium of early embryos (trophectoderm in mammals) can phagocytose defective pluripotent cells. Here, we explore the dynamic interactions between early embryos and bacteria. Quantitative live imaging of infection models developed in zebrafish embryos reveals the efficient phagocytic capability of surface epithelia in detecting, ingesting, and destroying infiltrated E. coli and S. aureus. In vivo single-cell interferences uncover actin-based epithelial zippering protrusions mediating bacterial phagocytosis, safeguarding developmental robustness upon infection. Transcriptomic and inter-scale dynamic analyses of phagocyte-bacteria interactions identify specific features of this epithelial phagocytic program. Notably, live imaging of mouse and human blastocysts supports a conserved role of the trophectoderm in bacterial phagocytosis. This defensive role of the surface epithelium against bacterial infection provides immunocompetence to early embryos, with relevant implications for understanding failures in human embryogenesis.We thank Andrés Hidalgo, Miguel Torres, and Stefanie Wculek for discussions and suggestions for the manuscript. We particularly thank Stephen Forrow, Barbara Tondelli, and Julien Colombelli from IRB, Tania García Becerra from IBMB, and Juana Fernández Rodríguez and Antoni Ventura from IDIBELL, for making possible the mouse experiments. We thank the support of the Animal facility, Bioimaging platform and Biostatistics Unit from IDIBELL, the Genomics unit from CRG, and the flow cytometry services from VHIR and the Parc Cientific of Barcelona. We thank Javier Moreno Santos for the mCherry-expressing strain, Senda Jimenez Delgado for sharing technical expertise, Nelly Henry for the bacteria with different motility, Jeff Rasmussen for the UAS-FYVE plasmid, Jan Roelof van Der Meer for the pHLuorin plasmids, Jeremy H. Lakey for the Caf1 plasmid, Shrikrishnan Sankaran for the fluorescent L. plantarum strains, Ákos T. Kovács for the fluorescent B. subtilis strains, Shang Yang for the BFP plasmid, Kayvon Pedram for the Rhobo6 dye, Verena Ruprecht for the dnFhod1 plasmid, Aitor Lavado for his assistance with the graphic design work, and Jordi Guiu, Victoriano Mulero, and Miguel Moreno Mateos for sharing resources. J.R.-C. acknowledges an FPU fellowship 22/04307 funded by Spanish Ministry of Science and Innovation and an FI fellowship 100057 from AGAUR. E.T. acknowledges funding from MCIN/AEI/10.13039/501100011033 “ERDF A way of making Europe” PID2021-125801OB-100 and AGAUR 2021SGR01545. A.V. acknowledges funding from Instituto de Salud Carlos III (PT23/00092). M.I. acknowledges funding from the Spanish Ministry of Science and Innovation (PID2020-115040GB-I00). E.H. acknowledges support from the Spanish Ministry of Science and Innovation (PID2020-117540GB-I00 and PID2023-151237NB-I00) and the AGAUR (2022-SGR-00974). E.H. held a Serra Hunter Lecturer position by the Generalitat of Catalonia at the University of Barcelona

    Women in crime: la representación de los personajes femeninos en la literatura negra contemporánea de Corea del Sur

    No full text
    Treball de Fi de Màster en Estudis d'Àsia-Pacífic en un context global. Curs 2024-2025Tutor: Carles Prado FontsLa novela negra es, por definición, un género literario que refleja la violencia más visceral presente en la sociedad que pretende reflejar. Si aplicamos esta idea a los personajes que la conforman, observamos cómo dicha violencia también los envuelve de diferentes maneras, empezando por su propia caracterización. En un género que ha alcanzado un notable éxito en Corea del Sur en los últimos años, se percibe que, desde su aparición en la península a inicios del siglo XX, ha existido un tratamiento diferenciado entre personajes masculinos y femeninos, a menudo definidos mediante características estereotipadas creadas por autores varones. Partiendo de estas premisas, el objetivo de este trabajo es analizar si la incorporación de escritoras ha implicado —o no— una evolución en la representación de los personajes femeninos. Para ello, se estudiarán tres obras escritas por mujeres a partir de 2010 —Lo que nunca sabrás (2015), El buen hijo (2019) e Hija única (2020)—, con la intención de comparar la representación femenina en estas novelas con la que se observa en las obras del siglo XX

    Maternal prenatal nut and seafood consumption and child neuropsychological function from 4 to 15 years of age: a population-based cohort study

    No full text
    Background: Understanding the role of maternal diet in early brain development is critical, as pregnancy represents a period of significant vulnerability and growth for the developing brain. Objectives: This study aims to assess the association between maternal nuts, total seafood, and large fatty fish consumption during pregnancy and offspring neuropsychological function ≤15 y, considering the potential mediation of omega-3 fatty acids. Methods: This study was part of The Spanish Childhood and Environment birth cohort, following 1737 mother-child pairs from pregnancy to age 15. Maternal diet was evaluated using a semiquantitative food frequency questionnaire, whereas children's neuropsychological function was measured through standardized computer-based tests. Attention (hit reaction time and its variability, HRT and HRT-SE) was measured with the Conners' Kiddie Continuous Performance Test and the Attention Network Test. Working memory (detectability in 2-back, d2', and 3-back tasks, d3') was evaluated using the N-back task. Fluid intelligence was assessed with Raven's Progressive Matrices and the Test of Primary Mental Abilities. Linear mixed-effects regression models assessed the association of nuts, seafood and large fatty fish with neuropsychological outcomes, whereas generalized structural equation modeling was used for mediation analyses. Results: Higher maternal nut consumption was significantly linked to improved attention [HRT-SE β = -0.05, 95% confidence interval (CI): -0.09, -0.00] and working memory (d2' β = 0.05, 95% CI: 0.00, 0.09, and d3' β = 0.06, 95% CI: 0.02, 0.11) in offspring. Greater consumption of large fatty fish was associated with better attention (HRT-SE β = -0.06, 95% CI: -0.10, -0.02; and HRT β = -0.04, 95% CI: -0.08, -0.00), and fluid intelligence (β = 0.08, 95% CI: 0.02, 0.13). Omega-3 fatty acids mediated 8%-14% of these effects on attention. Conclusions: Maternal diet at pregnancy and omega-3 intake may support long-term cognitive development in children and adolescents.This study was supported by Instituto de Salud Carlos III through the projects “CP14/00108, PI16/00261, and PI21/00266” (cofunded by the European Regional Development Fund “A way to make Europe”). The California Walnut Commission gave funding support. This study was also funded by other grants from Instituto de Salud Carlos III (Red INMA G03/176; CB06/02/0041; PI041436; PI081151 incl. FEDER funds; PI12/01890 incl. FEDER funds; CP13/00054 incl. FEDER funds; PI15/00118 incl. FEDER funds; CP16/00128 incl. FEDER funds; PI16/00118 incl. FEDER funds; PI16/00261 incl. FEDER funds; PI17/01194 incl. FEDER funds; PI17/01340 incl. FEDER funds; PI18/00547 incl. FEDER funds; PI20/01695 incl. FEDER funds), CIBERESP, Generalitat de Catalunya-CIRIT 1999SGR 00241, Generalitat de Catalunya-AGAUR (2009 SGR 501, 2014 SGR 822), Fundació La marató de TV3 (090430), Spanish Ministry of Economy and Competitiveness (SAF2012-32991 incl. FEDER funds), Agence Nationale de Securite Sanitaire de l’Alimentation de l’Environnement et du Travail (1262C0010; EST-2016 RF-21; EST-19 RF-04; 2019/1/233), EU Commission (261357, 308333, 603794; 634453; 825712 and 874583), FIS-PI042018 incl. FEDER funds, FIS-PI09/02311 incl. FEDER funds, FIS-PI13/02429 incl. FEDER funds, FIS-PI18/00909 incl. FEDER funds, CIBERESP, Obra Social Cajastur/Fundación Liberbank and UNIVERSIDAD DE OVIEDO, Department of Health of the Basque Government (2005111093), Provincial Government of Gipuzkoa (DFG06/002), and annual agreements with the municipalities of the study area (Zumarraga, Urretxu , Legazpi, Azkoitia y Azpeitia y Beasain). This study was also funded by grants from UE (FP7-ENV-2011 cod 282957, HEALTH.2010.2.4.5-1, cod 874583, and cod 101136566), Spain: ISCIII (G03/176; FIS-FEDER: PI03/1615, PI04/1509, PI06/1213, PI11/01007, PI11/02591, PI11/02038, PI12/00610, PI13/1944, PI13/2032, PI14/00891, PI14/01687, PI16/1288, PI17/00663, PI19/1338; P 23/1578), Miguel Servet-FEDER CP11/00178, CP15/00025, MSII16/00051, MS20/0006), Spanish Ministry of Universities (Margarita Salas Grant MS21-133, grant CAS21/00008), Generalitat Valenciana (CIAICO/2021/132, BEST/2020/059, AICO 2020/285, AICO/2021/182 and CIDEGENT/2019/064), Consejo General de Enfermería (PNI22_CGE45), FISABIO (UGP 15-230, UGP-15-244, UGP-15-249), and Alicia Koplowitz Foundation 2017. AP-M holds a predoctoral research training (PFIS) contract (grant FI22/00119) awarded by the Instituto de Salud Carlos III. JJ held a Miguel Servet-II contract since May 2023 (grant CPII19/00015) awarded by the Instituto de Salud Carlos III (cofounded by the European Social Fund “Investing in your future”). DH is supported by the LongITools project which has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 874739. The funders have no role in the study design, collection, management, analysis and interpretation of data, writing of the report or decision to submit it for publication

    Identification of novel compounds against Trypanosoma cruzi using AlphaFold structures

    No full text
    Chagas disease is a neglected tropical zoonosis caused by the protozoan Trypanosoma cruzi. The two approved medications for treating this disease show variable efficacy in the chronic phase, highlighting the need for new therapeutic interventions. This study explores a bioinformatics-driven approach to drug discovery using AlphaFold-predicted protein structures. Starting from a virtual screening of approximately 30,000 compounds, 24 were experimentally tested, and two already approved drugs, pimecrolimus and ledipasvir, demonstrated significant antiparasitic activity. These compounds were predicted to target previously uncharacterized T. cruzi proteins, ledipasvir interacting with a calpain-like protein and pimecrolimus likely binding a mitotic cyclin. Molecular dynamics simulations showed that pimecrolimus remains stable in the predicted binding site, while ledipasvir exhibits a higher RMSD. While experimental validation of these targets is needed, these findings underscore the potential of integrating AlphaFold structures into drug discovery strategies to accelerate the identification of new compounds against Chagas disease and other neglected tropical diseases. Summary: We performed a virtual screening experiment with T. cruzi AlphaFold protein models and a compound collection of more than 30,000 compounds. We tested the top ranked compounds in an in vitro setting, and found two promising candidates for drug repurposing against Chagas disease: pimecrolimus and ledipasvir

    0

    full texts

    43,569

    metadata records
    Updated in last 30 days.
    UPF Digital Repository
    Access Repository Dashboard
    Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇