http://hdl.handle.net/10366/159929 Sat, 18 Apr 2026 09:17:40 GMT 2026-04-18T09:17:40Z http://hdl.handle.net/10366/169824 [EN]Recent advances in machine learning and deep learning have significantly impacted multiple domains, including computervision, natural language processing and cybersecurity. In the context of increasingly sophisticated Advanced Persistent Threats(APTs), deep learning models have shown strong potential for network intrusion detection by addressing the limitations of tra-ditional methods. This study presents a comparative evaluation of classical and deep learning models for APT detection, high-lighting the ability of deep architectures, such as Convolutional Neural Networks and Long Short-Term Memory networks, toautomatically extract complex temporal and spatial patterns from network traffic data. A key objective is to maximise detectionaccuracy while minimising false positives and false negatives. Experimental results show that Convolutional Neural Networksapplied to the SCVIC-APT-2021 dataset achieved outstanding performance, with 99.24% accuracy, 99.39% precision, 99.24% re-call and a 99.24% F1-score. These results confirm the robustness of deep learning techniques for APT detection and underscoretheir effectiveness in identifying malicious activity in modern network environments. Tue, 09 Dec 2025 00:00:00 GMT http://hdl.handle.net/10366/169824 2025-12-09T00:00:00Z http://hdl.handle.net/10366/164740 [ES]Sumérgete en los vibrantes descubrimientos de las X Jornadas Internacionales de Enseñanza y Aprendizaje de la Estadística y la Investigación Operativa, celebradas en Salamanca en julio de 2023. Este libro recoge las contribuciones innovadoras y las reflexiones clave presentadas por académicos, profesionales y estudiantes de diversas instituciones y países. Desde talleres prácticos hasta charlas-coloquio, las jornadas destacaron la importancia de la innovación educativa en la formación estadística y operativa en todos los niveles educativos. Organizado por el Departamento de Estadística de la Universidad de Salamanca y el Grupo de Enseñanza y Aprendizaje de la Estadística y la Investigación Operativa (GENAEIO), perteneciente a la Sociedad Española de Estadística e Investigación Operativa (SEIO), este evento fue un hito en el campo, siendo el más multitudinario hasta la fecha y fomentando el intercambio de ideas y experiencias entre una comunidad global de educadores y expertos. Una lectura esencial para quienes buscan avanzar en la enseñanza y el aprendizaje de la Estadística y la Investigación Operativa, cruciales en el panorama académico actual. Mon, 01 Jul 2024 00:00:00 GMT http://hdl.handle.net/10366/164740 2024-07-01T00:00:00Z http://hdl.handle.net/10366/164739 [ES]Los ganaderos están interesados en conseguir mejores producciones de una manera eficiente y, por ello, siempre han tratado de innovar en la mejora genética de sus razas ganaderas mediante el uso de sementales testados mejorantes que ayuden a elevar los valores productivos sin necesidad de cruzamientos con otras razas foráneas. El manejo de los conocimientos en este campo es fundamental para los futuros profesionales técnicos en ganadería pues es interesante para la conservación de las razas y el patrimonio genético. El objetivo de este trabajo es mejorar la aptitud cárnica de los animales gracias a una mejor selección de los reproductores aplicando análisis estadísticos de las ganancias medias diarias o la prolificidad de las ovejas, con el fin de tomar las decisiones más adecuadas relacionadas con la producción ovina. Este estudio forma parte de en un proyecto de innovación educativa que pretende enseñar a los estudiantes del Grado Superior de Ganadería y Asistencia en Sanidad Animal (GASA) a realizar una investigación sobre la gestión de la producción ovina. En este trabajo participaron 60 alumnos de ambos turnos del módulo de Gestión de la Producción Animal del primer curso del ciclo formativo de Grado Superior de GASA del Instituto de Educación Secundaria San Isidro de Talavera de la Reina. Este centro docente cuenta con dos razas rústicas autóctonas de Castilla La Mancha: manchega y talaverana, sobre las que se realiza el estudio, estando en peligro de extinción la segunda y la variedad negra de la primera. Mon, 01 Jan 2024 00:00:00 GMT http://hdl.handle.net/10366/164739 2024-01-01T00:00:00Z http://hdl.handle.net/10366/164738 [ES]Las Tecnologías de la Información y la Comunicación (TIC) han revolucionado la forma en que los estudiantes adquieren conocimientos y habilidades. En concreto, el uso efectivo de las TIC en el aprendizaje de la Estadística puede ofrecer grandes beneficios en el contexto de la Educación Superior. El objetivo de este trabajo es conocer el uso de las TIC en el aprendizaje de la Estadística en la Educación Superior actual. Para ello, se aplicó un cuestionario a 78 estudiantes universitarios, 41% mujeres y 59% hombres, de la Escuela Politécnica Superior de Zamora (48,7%) y las facultades de Ciencias (25,6%), Ciencias Químicas (3,8%) y Ciencias Sociales (21,8%) de la Universidad de Salamanca. Los resultados mostraron actitudes positivas ante el uso de las TIC. La mayoría de los estudiantes indicaron que los profesores deben utilizar las TIC para mejorar la calidad de los procesos de aprendizaje, es imprescindible incorporarlas en las aulas universitarias y que proporcionan flexibilidad de espacio y tiempo para la comunicación entre los miembros de la comunidad educativa. No se observaron niveles de conocimiento sobre las tecnologías altos, los más elevados fueron sobre buscadores de información en red del tipo Google y los sistemas de comunicación (correo electrónico, foro, chat, videoconferencia, etc.) Sin embargo, destacaron niveles de conocimiento bajos de programas para el análisis de datos, como SPSS, o programas educativos de autor. Además, las herramientas de usuario y programas básicos del tipo Word, Power Point, etc. y los buscadores como Google fueron las tecnologías más utilizadas por los universitarios, mientras que para el aprendizaje de la Estadística los menos usados fueron los espacios de interacción social como Facebook, los programas para la edición de imagen, audio y vídeo, las plataformas virtuales de enseñanza-aprendizaje, como Moodle, y los materiales virtuales y recursos en red para la enseñanza-aprendizaje como portafolios electrónico, Wikis o videojuegos. Mon, 01 Jan 2024 00:00:00 GMT http://hdl.handle.net/10366/164738 2024-01-01T00:00:00Z http://hdl.handle.net/10366/164733 [EN]Vespa velutina, an invasive insect introduced to Europe from Asia, is the primary predator of honeybees, significantly contributing to the decline of their populations. Additionally, Vespa velutina has become a considerable threat to human health, as its sting can be lethal to individuals with allergies. The invasion of Vespa velutina disrupts ecosystems by threatening biodiversity and preventing pollination processes, and it also incurs socioeconomic costs, including negative impacts on apiculture and associated management expenses. To address these challenges, it is essential to develop fast and user-friendly automatic identification tools for Vespa velutina. This study proposes to design an artificial intelligence model capable of recognizing and identifying Vespa velutina among various insects. Such a model would enable the creation of devices that can automatically transmit images and geolocations in real-time, thereby enhancing the response efficiency of relevant authorities. The results of this work demonstrate the feasibility of accurately recognizing Vespa velutina using artificial intelligence technology, which supports the implementation of automated systems that slow the spread of this invasive species and protect the beekeeping ecosystem Mon, 01 Jan 2024 00:00:00 GMT http://hdl.handle.net/10366/164733 2024-01-01T00:00:00Z http://hdl.handle.net/10366/164732 [EN]The Internet of Things (IoT) is widely used in many applications including patient monitoring systems. The purpose of healthcare systems is to monitor the patient in order to prevent risks, deal with critical cases quickly, and establish long-distance communication for remote treatments. The IoT has a long-term impact on patient monitoring, patient management, patient physiological information, and critical care. The sensors are connected to the patient to collect the data which are first sent to system controls and then autonomously to healthcare providers. There are a variety of biosensors that send the medical information to mobile applications or websites via wireless network. Healthcare providers are thus enabled to monitor the patient and control the treatment outside of hospital walls. Therefore, the IoT medical devices require accurate patient monitoring methods in order to predict patient condition more precisely, and increase the efficiency of the network. An overview of patient monitoring systems based on machine learning in the IoT is provided in the following article. Wed, 01 May 2024 00:00:00 GMT http://hdl.handle.net/10366/164732 2024-05-01T00:00:00Z http://hdl.handle.net/10366/164690 [EN[The field of image recognition is extensively researched, with applications addressing numerous challenges posed by the scientific community. Notably among these challenges are those related to individual safety. This article presents a system designed for the application of image recognition in the realm of Occupational Risk Prevention—a concern of paramount importance due to the imperative of preventing workplace accidents as falls, collisions, or other types of accidents for the benefit of both workers and enterprises. In this study, convolutional neural networks are employed due to their exceptional efficacy in image recognition. Leveraging this technology, the focus is on the recognition of safety signs used in Occupational Risk Prevention. The primary objective is to enable the recognition of these signs regardless of their orientation or potential degradation, phenomena commonly observed due to regular exposure to environmental elements or deliberate defacement. The results of this research substantiate the feasibility of integrating this technology into devices capable of promptly alerting individuals to potential risks. However, to improve classification capabilities, especially for highly degraded or complex images, a larger and more diverse data set might be needed, including real-world images that introduce greater entropy and variability. Implementing such a system would provide workers and companies with a proactive measure against workplace accidents, thereby enhancing overall safety in occupational environments. Wed, 27 Nov 2024 00:00:00 GMT http://hdl.handle.net/10366/164690 2024-11-27T00:00:00Z http://hdl.handle.net/10366/164689 [EN]Readmissions are an indicator of hospital care quality; a high readmission rate is associated with adverse outcomes. This leads to an increase in healthcare costs and quality of life for patients. Developing predictive models for hospital readmissions provides opportunities to select treatments and implement preventive measures. The aim of this study is to develop predictive models for the readmission risk of patients with schizophrenia, combining the particle swarm optimization (PSO) algorithm with machine learning classification algorithms. The database used in the study includes a total of 6089 readmission records of patients with schizophrenia. These records were collected from 11 public hospitals in Castilla and León, Spain, in the period 2005–2015. The results of the study show that the Random Forest algorithm combined with PSO achieved the best results across the evaluated performance metrics: AUC = 0.860, recall = 0.959, accuracy = 0.844, and F1-score = 0.907. The development of these new models contributes to -improving patient care. Additionally, they enable preventive measures to reduce costs in healthcare systems. Wed, 11 Dec 2024 00:00:00 GMT http://hdl.handle.net/10366/164689 2024-12-11T00:00:00Z http://hdl.handle.net/10366/161428 [E]Earth observation satellites capture panchromatic images at high spatial resolution and multispectral images at lower resolution to optimize the use of their onboard energy sources. This results in a technical necessity to synthesize high-resolution multispectral images from these data. Pansharpening techniques aim to combine the spatial detail of panchromatic images with the spectral information of multispectral images. However, due to the discrete nature of these images and their varying local statistical properties, many pansharpening methods suffer from numerical artifacts such as chromatic and spatial distortions. This paper introduces the L0-Norm-based pansharpening method (L0pan), which addressed these challenges by maximizing the number of similar pixels between the synthesized pansharpened image and the original panchromatic and multispectral images. L0pan was optimized using a population-based colony search algorithm, enabling it to effectively balance both chromatic fidelity and spatial resolution. Extensive experiments across nine different datasets and comparison with nine other pansharpening methods using ten quality metrics demonstrated that L0pan significantly outperformed its counterparts. Notably, the colony search algorithm yielded the best overall results, highlighting the algorithm's strength in refining pansharpening accuracy. This study contributed to the advancement of pansharpening techniques, offering a method that preserved both chromatic and spatial details more effectively than existing approaches. Mon, 18 Nov 2024 00:00:00 GMT http://hdl.handle.net/10366/161428 2024-11-18T00:00:00Z http://hdl.handle.net/10366/161086 [EN]Color quantization ( CQ ) is a classical image processing operation that reduces the number of distinct colors in a given image. Although the idea of CQ dates back to the early 1970s, the first true CQ algorithm, median-cut, was proposed later in 1980. Since then, hundreds of publications have investigated the topic of CQ, proposing dozens of algorithms. A vast majority of these publications demonstrate their results on small datasets, containing a handful of images of mixed quality. Furthermore, the reproducibility of CQ research is often limited due to the use of private test images or public test images with multiple non-identical copies on the World Wide Web or restrictive licenses. To address these problems, we curated a large, diverse, and high-quality dataset of 24-bit color images called CQ 100 and released it under a permissive license. We present an overview of CQ 100 and demonstrate its use in comparing CQ algorithms. Wed, 07 Jun 2023 00:00:00 GMT http://hdl.handle.net/10366/161086 2023-06-07T00:00:00Z http://hdl.handle.net/10366/161085 [EN]The objective of the color quantization problem is to reduce the number of different colors of an image, in order to obtain a new image as similar as possible to the original. This is a complex problem and several solution techniques have been proposed to solve it. Among the most novel solution methods are those that apply swarm-based algorithms. These algorithms define an interesting solution approach, since they have been successfully applied to solve many different problems. This paper presents a color quantization method that combines the Artificial Bee Colony algorithm with the Ant-tree for Color Quantization algorithm, creating an improved version of a previous method that combines artificial bees with the K-means algorithm. Computational results show that the new method significantly reduces computing time compared to the initial method, and generates good quality images. Moreover, this new method generates better images than other well-known color quantization methods such as Wu’s method, Neuquant, Octree or the Variance-based method. Wed, 23 Jan 2019 00:00:00 GMT http://hdl.handle.net/10366/161085 2019-01-23T00:00:00Z http://hdl.handle.net/10366/161082 [EN]This article describes a color quantization algorithm that combines two swarm-based methods: Particle swarm optimization and artificial ants. The proposed method is based on a previous method that solves the quantization problem by combining the Particle swarm optimization algorithm with the K-means algorithm. K-means is a popular clustering method that has been applied to solve a variety of problems, including the color quantization problem. Nevertheless, it is a time-consuming method, which makes combining the Particle swarm optimization algorithm and K-means less suitable than other color quantization techniques. The proposed method, however, discards the K-means algorithm and applies the Ant-tree for color quantization algorithm in order to reduce execution time. This article shows that the new method outperforms the original one, since it requires less time to obtain higher quality images. In addition, the images produced are also of better quality than those produced by other well-known color quantization methods, such as Neuquant, Octree, Median-cut, Variance-based, Binary splitting and Wu’s methods. Wed, 01 Jan 2020 00:00:00 GMT http://hdl.handle.net/10366/161082 2020-01-01T00:00:00Z http://hdl.handle.net/10366/161047 [EN]A color quantization technique that combines the operations of two existing methods is proposed. The first method considered is the Greedy orthogonal bi-partitioning method. This is a very popular technique in the color quantization field that can obtain a solution quickly. The second method, called Ant-tree for color quantization, was recently proposed and can obtain better images than some other color quantization techniques. The solution described in this article combines both methods to obtain images with good quality at a low computational cost. The resulting images are always better than those generated by each method applied separately. In addition, the results also improve those obtained by other well-known color quantization methods, such as Octree, Median-cut, Neuquant, Binary splitting or Variance-based methods. The features of the proposed method make it suitable for real-time image processing applications, which are related to many practical problems in diverse disciplines, such as medicine and engineering. Tue, 01 Jan 2019 00:00:00 GMT http://hdl.handle.net/10366/161047 2019-01-01T00:00:00Z http://hdl.handle.net/10366/160754 [EN]Currently, web services-based applications have an important presence in public and private organizations. The vulnerabilities that these types of applications may have pose an inherent potential risk to the business model of these organizations. These applications have the inherent risk of being used by organizations in such a way that their activity is affected and they become the main entry point for attackers who want to breach their security. The main barrier to this type of attack are web application firewalls (WAF), which are responsible for processing Hypertext Transfer Protocol requests between clients and web servers, classifying them and rejecting malicious requests. This type of (WAF) applications, for the most part, have regular expressions that correspond to general rules and allow detecting malicious requests that follow a pattern contained in them. However, due to the knowledge of these rules by attackers, it is easy to circumvent security and to impersonate a malicious request by an innocuous request. Therefore, in this article, we present a study of different models based on artificial intelligence techniques as Naïve Bayes, k-nearest neighbors, support vector machines, and linear regression to test their effectiveness in detecting malicious requests from a synthetic dataset containing more than 100,000 requests. The results obtained show that the implementation of these methods optimize the detection of malicious requests obtaining results between 92% and 99% of success in their classification. Mon, 27 Nov 2023 00:00:00 GMT http://hdl.handle.net/10366/160754 2023-11-27T00:00:00Z http://hdl.handle.net/10366/160753 [EN]Introduction: Today’s society is aware that healthy aging favors quality of life in the future, even more so as life expectancy increases in populations such as Europe. As in countries such as Japan, it is necessary for institutions to provide social services to support the elderly, with the aim of achieving an optimal quality of life for these people. The aim of this study is to analyze the different types of social services and activities that certain institutions provide to the elderly in order to find areas for improvement or to propose relationships between them that will benefit both users and institutions. Methods: Official data from Junta de Castilla y León (Spain) on social services for the elderly in the 9 provinces of the autonomous community of Castilla y León from 2007 to 2021 were analysed using multivariate statistical techniques. Results: Throughout the period under analysis, there is an association between the number of places in public and private non-profit residential centers for the elderly and the number of places in day-care centers or the number of students in the Inter-University Experience Programme. The variables associated with the telecare programme are related to the number of people under guardianship. On the other hand, three well-differentiated clusters of provinces of Castilla y León were observed. Discussion: Our findings have implications for the quality of life of the elderly, as the differences in social services in the areas analysed have a direct impact on the health of the elderly. Fri, 24 Nov 2023 00:00:00 GMT http://hdl.handle.net/10366/160753 2023-11-24T00:00:00Z http://hdl.handle.net/10366/160752 [EN]This article analyzes various color quantization methods using multiple image quality assessment indices. Experiments were conducted with ten color quantization methods and eight image quality indices on a dataset containing 100 RGB color images. The set of color quantization methods selected for this study includes well-known methods used by many researchers as a baseline against which to compare new methods. On the other hand, the image quality assessment indices selected are the following: mean squared error, mean absolute error, peak signal-to-noise ratio, structural similarity index, multi-scale structural similarity index, visual information fidelity index, universal image quality index, and spectral angle mapper index. The selected indices not only include the most popular indices in the color quantization literature but also more recent ones that have not yet been adopted in the aforementioned literature. The analysis of the results indicates that the conventional assessment indices used in the color quantization literature generate different results from those obtained by newer indices that take into account the visual characteristics of the images. Therefore, when comparing color quantization methods, it is recommended not to use a single index based solely on pixelwise comparisons, as is the case with most studies to date, but rather to use several indices that consider the various characteristics of the human visual system. Thu, 25 Jan 2024 00:00:00 GMT http://hdl.handle.net/10366/160752 2024-01-25T00:00:00Z