39936 research outputs found
Sort by
Sistema de personalización de recomendaciones para eventos y etiqueta basado en el clima
No full textEste Trabajo de Fin de Grado tiene como objetivo el desarrollo de un sistema de recomendación de eventos culturales en la Comunidad de Madrid, adaptado a las preferencias individuales de los usuarios. Con el aumento de la oferta cultural en la ciudad y la creciente necesidad de soluciones que faciliten la selección de eventos, se propone una herramienta digital que permita personalizar las recomendaciones en función de variables como el precio, el horario, la temática y la ubicación.
El sistema utiliza técnicas de similitud para ajustar los eventos a las preferencias específicas del usuario, como el horario y el precio. Estas funciones permiten calcular de manera flexible la “proximidad” de un evento respecto a las preferencias del usuario, permitiendo incluir eventos que, aunque no cumplan con todos los requisitos al pie de la letra, son suficientemente cercanos en términos de los parámetros seleccionados. Además, el sistema integra datos contextuales en tiempo real, como las condiciones climáticas, que impactan directamente en la elección de los eventos, especialmente los de carácter al aire libre, lo que da un valor adicional al proyecto frente a sistemas tradicionales. Complementando esto, se utiliza un proceso de clasificación automática de temáticas mediante un modelo de clasificación zero-shot para mejorar la asignación de eventos según las características descritas.
A lo largo del proyecto, se analizan los retos técnicos y las soluciones adoptadas, con especial énfasis en la implementación, pruebas y validación del sistema. Se identifican también las limitaciones actuales, como el uso de archivos Excel para el almacenamiento de datos y la necesidad de una mayor escalabilidad, lo que se podrá abordar con el tiempo mediante el uso de bases de datos relacionales.
Este trabajo demuestra la viabilidad de un sistema innovador que facilita el acceso a eventos culturales personalizados con el objetivo de promover la inclusión y la redistribución de la oferta cultural en todos los distritos de la ciudad.Doble Grado en Ingeniería Informática y Administración de Empresa
A Low-cost Experimental Setup for Determination of the Residual Magnetic Dipole of Small Satellites
Full text linkThe residual magnetic dipole (RMD) moment is one of the primary disturbances in satellites, making its characterization essential for developing effective control strategies. This work presents a low-cost and compact experimental setup consisting of 12 cost-effective magnetic sensors, a nonmetallic structure, and dedicated software with the algorithms to estimate the dipole’s magnitude, direction, and location. In order to study the performance of the setup, three analyses were carried out. First, a simulation framework was prepared to study through Monte Carlo analysis the role of the ambient magnetic field noise, sensor installation errors, and number of sensors. It was found that, under typical laboratory conditions and installation errors, at least eight sensors should be used to obtain reliable estimations. Moreover, the simulation revealed that, in an environment with 150 nT of ambient noise and installation accuracy better than 0.1 mm and 0.1°, the errors of the estimated dipole magnitude, direction, and location are below 4%, 6°, and 7 mm for dipoles above 0.01 A ⋅ m2. Second, the performance of the experimental setup was tested by using a dipole generator. In an environment with 450 nT of ambient noise and for moderate fabrication quality, the setup achieved an error smaller than 11% in magnitude, 10° in direction, and 15 mm in location for dipoles greater than 0.2 A ⋅ m2. Finally, the setup was used to characterize the dipole parameters of a deorbit device (DD) based on electrodynamic tether technology. The results were consistent with an independent characterization performed with a conventional Helmholtz cage-based method.This work was supported by European Union’s Horizon Europe Research and Innovation Program through the E.T.PACK-F Project under Grant 10105816
La dimensión transnacional del cine español contemporáneo a través de las estrategias de producción cinematográfica de Telecinco Cinema
Full text linkCon el transcurso de las tres últimas décadas, Telecinco Cinema ha logrado erigirse como una de las productoras más relevantes a la hora de respaldar los estrenos comerciales de mayor recaudación en España. Entre las casi 90 películas que han conformado su trayectoria, destacan producciones muy relevantes de manos de los cineastas más populares, como Álex de la Iglesia, Alejandro Amenábar o Juan Antonio Bayona. No obstante, Telecinco Cinema parte de un punto de origen que supuso un hito en la historia de las industrias cinematográfica y televisiva españolas como es el paisaje audiovisual surgido a partir de las transposiciones de la Directiva de "Televisión Sin Fronteras" (89/552/CEE) en 1999. A partir de entonces, se impuso un 5% de los ingresos brutos a los operadores de televisión la participación financiera en la producción cinematográfica nacional y europea
PIC/Fluid simulations of the plasma expansion in a planar magnetic arch
Full text linkMagnetic arches (MA) (i.e. the magnetic topology that emerges when placing two magnetic nozzles with opposite polarities side by side) are an attractive option for the clustering of multiple electrodeless plasma thrusters, as they are characterized by a zero magnetic dipole moment and thus allow a reduction of perturbing magnetic forces on the spacecraft. This work employs the hybrid code EP2PLUS to simulate and study the plasma expansion for such a magnetic topology in the planar limit. First, a reference simulation is used to analyze the leading physical mechanisms that govern the plume properties. Ions are thus found to be characterized by a double peaked velocity distribution function close to the symmetry plane, where the plasma beams emitted by the two thrusters merge, while the magnetic force acting on electrons is shown to shape both the lateral confinement of the plume, and the thrust profile provided. Second, a parametric sweep on the strength of the magnetic field shows that its influence on the propulsive properties and on the characteristics of the plume saturates for values of the Hall parameter larger than around 10. Beyond this value of the Hall parameter, only the in-plane electron currents are found to be particularly sensitive both to the magnetization levels and the boundary conditions employed, although they are also largely decoupled from the rest of plasma properties. Finally, background pressure effects were considered by including collisions with neutral atoms in the simulations, highlighting the relevance of neutral entrainment in the modification of the plume properties and in the propulsive performance of the MA.The authors would like to thank Celian Boyé for his contributions in the initial definition of the simulation setup. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 950466)
Assessing ChatGPT's ability to detect and correct programming errors in stata do-files
No full textThis paper evaluates the efectiveness of ChatGPT in debugging Stata, a proprietary econometric software, with a focus on accuracy, hallucination tendencies, and the presence of illusory expertise. By systematically testing ChatGPT-3.5-turbo and ChatGPT-4o across diferent error types and user expertise levels, the study finds that while ChatGPT signifcantly improves debugging performance in open-book mode, its accuracy remainshighly dependent on error complexity. Hallucinations, where the model generates plausible but incorrect error explanations, are particularly frequent in syntax errors and persist even with access to error logs. Advanced scripts present the greatest challenge, with ChatGPT- 4o achieving only a 35.7 percent success rate and exhibiting a 25 percent hallucination rate despite access to error messages. The findings align with broader research on largelanguage models in coding assistance, demonstrating that ChatGPT struggles with complex debugging tasks in proprietary software environments like Stata, where integration is restricted.Ricardo Mora acknowledges the fi nancial support by MICIU/AEI/10.13039/501100011033 grants
PID2023-146391NB-I00 and CEX2021-001181-M; and Comunidad de Madrid, grant EPUC3M11 (VPRICIT)
Obra española en servicios de vídeo bajo demanda por suscripción: disponibilidad y prominencia. Edición 2025
No full textInforme elaborado por el Grupo de Investigación Diversidad Audiovisual de la Universidad Carlos III de Madrid con el apoyo del Instituto Universitario del Cine Español (IUCE-UC3M)Este trabajo analiza la presencia de obra española en seis servicios de vídeo bajo demanda por suscripción (SVOD) en el mercado español. El informe se focaliza, por orden de llegada a España, en Netflix, Prime Video, Max, Apple TV+, Disney+ y SkyShowtime. Además de la presencia de obra española, se examinan los principales mecanismos utilizados para darle prominencia en los servicios. La disponibilidad de títulos españoles ha ido en aumento desde el año 2022, cuando comenzó a elaborarse este informe anual. La conclusión principal del informe es que todos los servicios nutren sus respectivos catálogos de obras españolas, aunque la cantidad y características de estas varían considerablemente. Incluyendo a las coproducciones internacionales, en su conjunto los seis catálogos ofertan 1.802 títulos españoles únicos, cifra que implica un incremento de 741 obras respecto a enero de 2022 (+69,8%). Prime Video, con 1.032 títulos, es el servicio que dispone de una mayor cantidad de contenidos españoles; seguido por Netflix, que ofrece 674 obras españolas. Al poner estos datos en relación con el tamaño total de cada catálogo, SkyShowtime es el servicio que dispone de un mayor porcentaje de obra local, pues los títulos españoles representan un 15,8% de su catálogo. Prime Video se sitúa cerca, con una oferta de obras constituida en un 15% por títulos españoles. En Netflix y Max este porcentaje se ubica alrededor del 8%, mientras que en Disney+ y Apple TV+ representan menos de un 3% en cada caso. La consolidación de tendencias en relación con los contenidos ofertados y su forma de comercialización, hacen de las interfaces de los servicios entornos crecientemente complejos de analizar. Los servicios estudiados utilizan diferentes mecanismos para otorgar prominencia a la obra española de sus catálogos. Con la excepción de Apple TV+, todos disponen de secciones dedicadas a las obras españolas. Asimismo, los buscadores de cada servicio varían en el número y tipo de resultados que arrojan al introducir palabras clave relacionadas con la obra local, siendo Netflix el que en mayor medida facilita la posibilidad de encontrar películas, series y otros tipos de contenidos españoles.Este informe fue concebido en el marco del proyecto de investigación "Diversidad y servicios audiovisuales bajo demanda por suscripción (PID2019-109639RB-I00)", financiado por el Ministerio de Ciencia, Innovación y Universidades y la Agencia Estatal de Investigación (MCIN/AEI/10.13039/501100011033/)
40 Years of Empirical Evidence of Cointegration and Nonlinear Equilibrium Correction in UK Money Demand since the XIX Century
No full textSince the influential works of Friedman and Schwartz (1963, 1982) and Hendry and Ericsson (1991), on the monetary history of the United States of America and the United Kingdom from 1876 to 1975, there has been a great concern in the literature about the instability of money demand functions. This concern together with the results of the New Keynesian models, produced the abandonment of money as an instrument of monetary policy. Recently, using M1 as the measure of money, Benati, Lucas, Nicolini and Weber (2021) have shown, for a shorter and recent period of time, that there is a stable long-run money demand for a long list of countries. However, to date there are no studies showing that alternative stable longrun and short-run money demand equations exist since the XIX century. By means of nonlinear cointegration and nonlinear equilibrium corrections (NEC), we present empirical evidence of stable nonlinear UK money demands models of real broad money balances from 1877 to 2023. The properties of these NEC models are assessed via Monte Carlo simulations. Rational polynomials error-correction models are used to generate a simple nonlinear Granger´s representation theorem together with a two-step estimation procedure, which satisfies well-stablished asymptotic conditions. As a byproduct, with four different but stable money demand specifications, we empirically identify key abrupt historical periods, corresponding to World Wars I and II, regulatory changes and the COVID period, generating a common 6.5% excess inflation effect, over the historical 2.2% constant average inflation rate since 1877.The corresponding author is A. Escribano ([email protected]), and he acknowledges the funding received from grants PID2022-151414OB-100 and
CEX2021-001181_M funded by MICIU/ AEI/10.13039/501100011033
Balancing generosity with profitability: The role of relative market price and value perceptions in crypto philanthropy
Full text linkConsidering the exponential increases in donations made in cryptocurrencies in recent years, the decision-making process that underlies crypto philanthropy demands deeper insights. Given that both individual donors and recipients are motivated to own multiple types of cryptocurrencies to strategically diversify their investment, our research examines the influence of different cryptocurrencies" relative market price on the donation amount. The results derived from on-chain data and five experiments indicate that donors tend to set a lower donation amount for higher-priced cryptocurrency than for lower-priced cryptocurrency to protect the profitability of their crypto portfolios. To alleviate the profitability concerns and encourage larger donations in higher-priced cryptocurrencies, potential recipients could intensify donation campaigns at times when analysts predict lower future values, or emphasize the number of cryptocurrency coins that would remain in the donor"s wallet after donation. By uncovering new insights into the effects of relative market prices, future value predictions, and crypto-wallet designs, our findings provide guidance for marketers, practitioners, and policymakers interested in crypto philanthropy.We are grateful for the research grants received from the Community of Madrid (2022-T1/SOC-24048), the Ramon Areces Foundation (CISP20S12340), Instituto Nacional de Ciberseguridad (INCIBE-CYBERTHREAT C115/23), and Agencia Estatal de Investigación (TED2021-129861B-I00)
Emotion forecasting: A transformer-based approach
No full textBackground:
Monitoring the emotional states of patients with psychiatric problems has always been challenging due to the noncontinuous nature of clinical assessments, the effect of the health care environment, and the inherent subjectivity of evaluation instruments. However, mental states in psychiatric disorders exhibit substantial variability over time, making real-time monitoring crucial for preventing risky situations and ensuring appropriate treatment.
Objective:
This study aimed to leverage new technologies and deep learning techniques to enable more objective, real-time monitoring of patients. This was achieved by passively monitoring variables such as step count, patient location, and sleep patterns using mobile devices. We aimed to predict patient self-reports and detect sudden variations in their emotional valence, identifying situations that may require clinical intervention.
Methods:
Data for this project were collected using the Evidence-Based Behavior (eB2) app, which records both passive and self-reported variables daily. Passive data refer to behavioral information gathered via the eB2 app through sensors embedded in mobile devices and wearables. These data were obtained from studies conducted in collaboration with hospitals and clinics that used eB2. We used hidden Markov models (HMMs) to address missing data and transformer deep neural networks for time-series forecasting. Finally, classification algorithms were applied to predict several variables, including emotional state and responses to the Patient Health Questionnaire-9.
Results:
Through real-time patient monitoring, we demonstrated the ability to accurately predict patients" emotional states and anticipate changes over time. Specifically, our approach achieved high accuracy (0.93) and a receiver operating characteristic (ROC) area under the curve (AUC) of 0.98 for emotional valence classification. For predicting emotional state changes 1 day in advance, we obtained an ROC AUC of 0.87. Furthermore, we demonstrated the feasibility of forecasting responses to the Patient Health Questionnaire-9, with particularly strong performance for certain questions. For example, in question 9, related to suicidal ideation, our model achieved an accuracy of 0.9 and an ROC AUC of 0.77 for predicting the next day"s response. Moreover, we illustrated the enhanced stability of multivariate time-series forecasting when HMM preprocessing was combined with a transformer model, as opposed to other time-series forecasting methods, such as recurrent neural networks or long short-term memory cells.
Conclusions:
The stability of multivariate time-series forecasting improved when HMM preprocessing was combined with a transformer model, as opposed to other time-series forecasting methods (eg, recurrent neural network and long short-term memory), leveraging the attention mechanisms to capture longer time dependencies and gain interpretability. We showed the potential to assess the emotional state of a patient and the scores of psychiatric questionnaires from passive variables in advance. This allows real-time monitoring of patients and hence better risk detection and treatment adjustment.LP-A was supported by the Spanish Instituto de Salud Carlos III (PMP22/00032) and by the Spanish Ministry of Science, Innovation, and Universities through a doctoral FPU grant (Formación de Profesorado Universitario; FPU23/03723). AA-R was partially supported by the Spanish Instituto de Salud Carlos III (PMP22/00032); Ministerio de Ciencia, Innovación y Universidades (CPP2022- 009537); and Ministerio de Ciencia e Innovación jointly with the European Commission (ERDF, European Regional Development Fund; PID2021-123182OB-I00 and PID2021-125159NB-I00). DR was partially supported by the Office of Naval Research Global (N62909-23-1-2002), the Comunidad de Madrid (IRIS: Hacia un asistente personal para la mejora del bienestar emocional. Ayudas para la realización de Doctorados Industriales en la Comunidad de Madrid 2023; IND2023/TIC-27508), the Spanish Ministry of Economic Affairs and Digital Transformation, and the European Union-NextGenerationEU through the UNICO 5G I+D SORUS project. PMO was supported by the Comunidad de Madrid (IND2022/TIC-23550), ELLIS Unit Madrid (European Laboratory for Learning and Intelligent Systems), and 2024 Leonardo Grant for Scientific Research and Cultural Creation from the BBVA Foundation (Banco Bilbao Vizcaya Argentaria). Both DR and PMO were supported by MICIU/AEI/10.13039/501100011033/FEDER, UE (PID2021-123182OB-I00; EPiCENTER). This work was also supported by the Comunidad de Madrid through the IDEA-CM project (TEC-2024/COM-89)
Apuntes de control de sistemas II
Full text linkUno de los problemas con los que tradicionalmente se tienen que enfrentar aquellos que trabajan en el control o en la identificación de procesos estriba en la existencia de una diversidad de herramientas matemáticas para estudiar, analizar y diseñar dichos sistemas