Repositorio Institucional
El repositorio institucional recoge la producción científica del personal docente e investigador de la Universidad de Deusto. Su propósito es reunir, archivar, preservar y aumentar la visibilidad en acceso abierto de los resultados de investigación.
Comunidades en DSpace
Selecciona una comunidad para explorar sus colecciones.
Envíos recientes
Electricity forecast adapted to ocean conditions: the Mutriku case study
(Elsevier Ltd, 2024-08) Casas, Isabel; Lekube, Jon
It is essential for any wave power plant at the commercialising stage to accurately model and forecast its production to ensure a good functioning and an efficient trading in the daily wholesale electricity market. Altogether, a reliable forecast is a required step for the establishment of wave energy as a sustainable and viable source of electricity. We propose models whose coefficients adapt to the variation of ocean conditions and apply them to model/estimate and forecast the electricity production of two oscillating wave converters from the Mutriku Wave Power Plant during 2018. After a statistical analysis comparing different approaches, we recommend to model the production using a regression with coefficients depending on the wave height and to do short-term forecast using an autoregressive function with coefficients also depending on the wave height.
Exercise addiction and intimate partner violence: the role of impulsivity, self-esteem, and emotional dependence
(Multidisciplinary Digital Publishing Institute (MDPI), 2024-05) Olave, Leticia; Iruarrizaga, Itziar; Herrero Lázaro, Marta; Macía Guerrero, Patricia; Momeñe López, Janire; Macía Guerrero, Laura; Muñiz Casado, José Antonio; Estévez Gutiérrez, Ana
Given the scarcity of studies linking exercise addiction to intimate partner violence, the present study aims to analyze the relationship between these variables and examine the potential mediating roles of emotional dependence, impulsivity, and self-esteem. This is a non-experimental, cross-sectional correlational design study. The sample comprised 887 university students (86% women, mean age 20.82 years, SD = 3.63). Elevated levels of exercise addiction were associated with increased impulsivity, emotional dependence, and exerted violence, as well as decreased self-esteem and perceived violence. Mediation models were tested, explaining 7% of the variance in received violence, 13% of the variance in exerted violence, and 6% of the variance in perceived violence. Higher levels of exercise addiction were linked to increased received and exerted violence and decreased perceived violence, attributed to the positive impact of exercise addiction on emotional dependence. This study highlights the mediating roles of self-esteem and impulsivity in the relationship between exercise addiction and partner violence. Identifying risk or vulnerability factors such as emotional dependence, impulsivity, and self-esteem related to exercise addiction and interpersonal partner violence is especially relevant for designing and implementing preventive interventions in the general young population.
Talent management digitalization and company size as a catalyst
(Multidisciplinary Digital Publishing Institute (MDPI), 2024-05) Martínez Morán, Pedro César; Díez Ruiz, Fernando; Solabarrieta, Josu; Fernández Rico Urgoiti, José María; Igoa Iraola, Elene
As companies increasingly undergo digital transformation, the role of talent management processes becomes pivotal in enhancing overall organizational performance. The objective of this research is to assess the extent to which greater digitalization in the talent management process is linked to company size. The research has addressed four research questions in order to explore the significance of talent management in corporate digital transformation, examining whether variations in the digitalization of these processes can be attributed to company size. A qualitative approach was employed, utilizing a questionnaire, and collecting responses from 202 organizations across diverse sectors. The findings reveal disparities in digitalization throughout the talent management process, with pronounced presence in the attracting, selecting, and rewarding phases, but diminishing in deployment and development, and further declining in planning. A positive correlation between company size and the adoption of specific digital platforms was observed. Larger enterprises exhibit greater utilization of digital platforms in talent deployment and development. Moreover, corporate communication tools are consistently utilized in the rewarding phase, irrespective of company size. These findings offer practical insights for organizations aiming to optimize their digitalization strategies based on their scale, thereby contributing to more effective and tailored digitalization endeavours. The uniqueness of this research lies in its exploration of the influence of company size on the digitalization of talent management processes and its potential to explain variations across different stages of these processes.
Canicross pilot programme: basic considerations for its implementation as an extracurricular sports activity
(Multidisciplinary Digital Publishing Institute (MDPI), 2024-05) González Santamaría, Xabier; Borrajo, Erika; Sánchez Mencia, Eneko; Aurrekoetxea Casaus, Maite
Canicross is a sport that consists of running while being pulled by a dog in a natural environment. Due to the benefits to health and well-being that it brings to the people and dogs that practise it, this sport could be implemented as an animal-assisted intervention (AAI) in the educational field. Against this background, the present work describes a pilot experience carried out in an educational centre in the Basque Country for the implementation of canicross as an extracurricular sports activity. The main objective is to describe the educational potential of canicross as an extracurricular activity that fosters students’ values of respect for animals and physical activity (PA) in natural environments. The results of this innovative experience provide the first evidence of the valuable role of animal activities in this educational context, where current academic research is practically non-existent. In conclusion, we highlight the novelty of the proposal and the motivating effect that the dogs in this case had among the students in encouraging them to practise PA and adhere to values of respect and animal welfare.
A new quantum circuits of quantum convolutional neural network for X-ray images classification
(Institute of Electrical and Electronics Engineers Inc., 2024) Yousif, Mohammed; Al-Khateeb, Belal; García-Zapirain, Begoña
A common model for classifying images is the convolutional neural network (CNN), which has the benefit of effectively using data correlation information. Despite their remarkable success, classical CNNs may face challenges in achieving further improvements in accuracy, computational efficiency, explainability, and generalization. However, if the specified data dimension or model grows too large, CNN becomes difficult to train effectively with a slowdown processing. In order to address a problem using CNN utilizing quantum computing, Quantum Convolutional Neural Network (QCNN) proposes a novel quantum solution or enhances the functionality of an existing learning model in terms of processing time during training. This paper presents a comparative analysis between classical Convolutional Neural Networks (CNNs) and a novel quantum circuit architecture tailored for image-based tasks, emphasizing the adaptability and versatility of quantum circuits in enhancing feature extraction capabilities and then final accuracy and processing time. A MNIST and covidx-cxr3 datasets was used to train quantum-CNN models, and the results of these comparisons were made with traditional CNN performance. The results demonstrate that the suggested QCNN beat the traditional CNN in terms of recognition accuracy and processing speed (process time) when combined with cutting-edge feature extraction techniques. This superiority is particularly evident when trained on the covidx-cxr3 dataset, highlighting the potential for quantum computing to revolutionize image classification tasks.