This research explores the application of possibility theory in the stock market, aiming to improve trading outcomes by integrating various technical indicators. The study focuses on the Relative Strength Index (RSI), Moving Average Convergence/Divergence (MACD), and Moving Average Crossover (MAC) to generate actionable buy, sell, or hold signals. By using possibility theory to combine these indicators, the research seeks to provide a more accurate assessment than individual indicators alone. Results indicate that the possibility theory-based approach significantly enhances returns while mitigating risk compared to traditional methods.

This study examines the training needs in digital technologies of companies in the Valencian Community, in the context of the growing importance of digital transformation. Using a mixed methodology combining in-depth interviews and surveys, the perceptions, obstacles and training demands of more than 100 innovation-intensive companies with a track record of more than 5 years were analysed. The results reveal a significant gap between awareness of the importance of digitalisation and the eKective implementation of digital strategies. Participating companies demand practical, contextualised and management-led training, with a preference for blended learning and short durations. The study provides recommendations to adapt the training oKer to the specific needs of the Valencian business fabric, thus contributing to the literature on digital transformation in regional contexts.

The global climate crisis has accelerated the search for sustainable energy alternatives, positioning green hydrogen as a promising solution for the decarbonisation of various industrial sectors. This article takes a comprehensive look at the role of green hydrogen in Chile's energy transition, analysing its potential to reduce greenhouse gas emissions, promote sustainable economic development and position the country as a world leader in clean energy. It assesses the World Bank's recent investment of US$150 million in the 'Green Hydrogen Facility in Chile' programme, exploring its implications for the Chilean energy industry and the Latin American and Caribbean region. The study addresses the technical, economic, political and social aspects of green hydrogen development in Chile, including technological challenges, market opportunities, necessary public policies and expected socioeconomic impacts. It analyses the current energy context in Chile, characterised by an increasing share of renewable energies, and projects the country's potential to become a leading exporter of green hydrogen globally. It also examines the role of green hydrogen in the decarbonisation of key sectors such as mining, transport and heavy industry, as well as its potential to drive the creation of green jobs and the development of new value chains. The article concludes with a critical assessment of the prospects for green hydrogen in Chile, offering recommendations to maximise its benefits and mitigate the risks associated with its large-scale implementation.

The integration of deep learning into signal processing has changed how we decode & classify signals. This applies to areas, including neuroscience and wireless communication systems. In this paper, we together three new research studies that use deep learning techniques to improve signal processing. First, we look at convolutional neural networks (s). These are for decoding electrooculogram (EEG and local field potential (L) signals. The results show much better accuracy and efficiency in brain-computer interface (BCI) systems. Also, using these networks on IBM's TrueNorth neuromorphic chip shows it is possible to do real-time, exceptionally low power signal processing for neurobionics. Next, the paper reviews deep neural networks (DNNs) for channel decoding in communication systems. This study shows DNNs can reach near-optimal decoding performance, especially when using structured codes like polar codes. This study shows DNNs can reach near-optimal decoding performance, especially when using structured codes like polar codes. The networks can even do well with data they have not seen before. But there are some issues too. The curse of dimensionality and the difficulty of scaling these networks for longer codewords are important challenges discussed. Finally, we examine an adaptive ensemble deep learning model made for signal detection in Orthogonal Frequency Division Multiplexing (OFDM) systems. This model tackles big challenges in channel estimation and compensation by adjusting to changing conditions of multi-path channels. Because of this adjustment, there are major improvements in signal reliability & fewer symbol error rates (SER). Through these advanced methods, this paper highlights how deep learning impacts signal processing. It opens doors for future advancements in both neuroscience and wireless communication fields.

Indoor localization within buildings is paramount due to its diverse applications, including the Internet of Things (IoT), healthcare, and personnel monitoring. To improve the performance of indoor localization, this study proposes an algorithm combining Deep Neural Networks (DNNs) and Convolutional Neural Networks (CNNs) used within a hybrid model that uses both cellular technology, i.e, base station (BS) and Wi-Fi access point (AP). Within a realistic environment, we evaluate the performance of our approach across four scenarios: (1) (2 APs + 2 BSs), (2) (1 AP + 2 BSs), (3) (2 APs), and (4) (2 BSs). Additionally, we consider both stationary and seated user positions. Our results demonstrate significant improvements over existing methods. Specifically, in the 2 APs + 2 BSs scenario, we achieve an average Euclidean distance error of 1.04 meters, with a maximum error of 2.51 meters, and the Root Mean Square Error (RMSE) of 1.15 meters. We also obtain the corresponding Cumulative Distribution Function (CDF) for the error which indicates that 90% of the time, the error is less than 1.69 meters. Our hybrid model achieves over 99% accuracy in classifying different building floors and over 88% accuracy in identifying user states (e.g., standing or sitting). We validate our approach using smartphone-based indoor localization in a two story residential building with robust construction materials and double-layered walls.

How to build the arithmetic finding 3-uncolorable vertices in polynomial time in 3-coloring is studied. For planar graph G[V,E] the diagrammatic arithmetic computing jammed configurationwith the independent clip (DJC) is built in the color space according to the Grötzseh theorem. Each DJC corresponds a triplet in G[V,E] , which is the invariant set of three sets of monochromatic vertices with different colors under coloring transformation. The algorithmcancomplete the computation to found all DJC in 8 O(V ) . The vertex adjoins vertices of three monochromatic sets is 3-uncolorable. They are only the part of 3-uncolorable vertices. So, 3-coloring has 3-uncolorable vertices can be recognized in polynomial time.

During quantum era, new rules governing physical phenomena and reality contribute to the development of quantum technology, which will change people’s lives, in the near future. A great range of quantum technologies, based on quantum information theory and quantum computing, lead to new applications. Revolutionary quantum computing algorithms benefit from basic quantum mechanics properties, such as entanglement and superposition effects, to advance calculations in polynomial time. In essence, Quantum Mechanics (QM) is considered a cross – disciplinary STEM Physics field, aiming to advance the philosophy of Quantum Literacy (QL), which addresses the transdisciplinary nature of real world complex problems. QL addresses the challenges of quantum technologies learning and skills acquisition, through gamification and computing problem – solving environments. Simulations, virtual labs and interactive tools through serious games, create an attractive learning environment and promote at the same time an interdisciplinary field of physics, technology and information sciences. Technological interventions, often combined with contemporary teaching methods, provide guidelines for better and long-lasting learning outcomes. To this end, literature often emphasizes students’ engagement with serious games, especially if the content is interdisciplinary and lies in the crossroad section of Physics, Mathematics, Computer and Computational Sciences. Our work focuses on mapping the State of the Art among contemporary teaching methods and gamification techniques and tools, as applied for quantum mechanics concepts and quantum computing problems, to best advance QL. We also discuss the trends among contemporary teaching methods, effective gamification tools and quantum learning topics, fostering quantum physics knowledge in K-12 educational level.

Optimal current setpoints in the form of Maximum torque per ampere (MTPA) and Field Weakening (FW) curves are well-known for three-phase machines. Fivephase machines have more freedom in searching for optimum since the current waveform contains except the first harmonic also the third harmonic, which has a role in torque enhancement and waveform shaping. The problem does not allow for an analytical solution and has to be solved numerically. In this contribution, we analyze the resulting solution and provide guidance for its implementation. Specifically, we show that the optimal operation has MTPA and FW regimes that have an inner structure with multiple sub-regimes. This variety comes from the limits imposed on the maximum of the waveforms of the phase voltage and phase current (peak shaving). A taxonomy of these regimes is provided, and their mutual relations are demonstrated on two exemplary machines. Moreover, we show that commonly used assumptions on the symmetry of the waveforms yield suboptimal results. The theory is validated experimentally on a laboratory prototype of a fivephase PMSM drive of 3 kW.

An approach for calculation of the magnetic force arising between two electric current-carrying filaments having a circular and closed-curve of arbitrary shape is developed. The developed approach is based on the recently formulated segmentation method applied for the calculation of the mutual induction for a similar filament system. Employing the fact that any curve can be interpolated by a set of line segments with the desired accuracy and deriving the set of formulas for calculating of the magnetic force between a circular filament and line segment, the developed approach was also successfully applied for the estimation of the distribution of magnetic force along the closed-curve in addition to the resulting one. As illustrative examples, the calculation of the magnetic force and its distribution between the circular filament and the following closed-curves such as polygons, circles and a 3D curve was efficiently performed by using the developed approach. Also, the developed method was applied for the calculation of the resultant magnetic force between the rigid bodies including permanent magnets and current-carrying coils. The results of calculation were validated successfully by using FEM method and the analytical formulas available in the literature.

As artificial intelligence (AI) continues to advance, its application in healthcare is rapidly evolving from automating routine tasks to handling complex medical decision-making. This paper critically evaluates OpenAI's GPT-o1 model, a large language model, as a potential AI-powered medical practitioner. The study analyzes its performance across various domains of medical practice, including clinical knowledge, differential diagnosis, pharmacology, treatment planning, ethical decision-making, and patient communication. A diverse range of clinical scenarios-spanning cardiology, endocrinology, pulmonology, oncology, and other specialties-was used to assess the model's accuracy, clinical reasoning, and ability to adhere to current evidence-based guidelines. The results revealed that GPT-o1 performs impressively in areas such as diagnosing acute conditions, with an accuracy rate of over 95% in certain cases like STEMI, and offering evidence-based treatment plans, especially in chronic disease management. The model demonstrated significant ethical sensitivity, particularly in handling patient autonomy and end-of-life care, with more than 90% of its responses aligning with ethical best practices. However, challenges remain in the model's ability to manage complex, multifactorial cases that require nuanced reasoning and contextual awareness. In scenarios involving differential diagnoses, such as tuberculosis in endemic regions, GPT-o1's accuracy dropped by 20% due to a lack of consideration of broader epidemiological factors. Moreover, the study found that while GPT-o1 excels in providing medically sound responses that adhere to clinical guidelines, it struggles in emergent, time-sensitive situations where rapid decision-making is critical. The model's response time in medical emergencies, like diabetic ketoacidosis (DKA) or hyperosmolar hyperglycemic state (HHS), lagged behind human judgment by 15-20%, raising concerns about its viability in high-pressure environments.Despite its high overall performance, with accuracy rates ranging between 85-95% across most specialties, GPT-o1 still requires substantial development to fully replicate the nuanced decision-making of human practitioners. This study concludes that while AI holds enormous potential to support healthcare providers, particularly in diagnostics and treatment planning, its current limitations emphasize the need for continued human oversight. To integrate AI models like GPT-o1 into healthcare systems effectively, ongoing collaboration between AI developers and medical professionals is crucial to ensure safety, reliability, and the continuous incorporation of the latest medical research.

This comprehensive review critically evaluates the evolving role of hybrid AC/DC power systems in modern energy landscapes, emphasizing their pivotal function in integrating both renewable and conventional power sources. As the integration of renewable resources scales up, optimizing operational efficiency and stability becomes increasingly challenging. This review delves into the advanced optimal power flow (OPF) strategies that not only improve load distribution and energy efficiency but also enhance the reliability and resilience of power grids. By exploring various OPF methodologies, including conventional, probabilistic, and security-constrained OPF, this study underscores the importance of innovative approaches to ensure economic, safe, and reliable grid operations under complex and dynamic energy demands.

Head injuries pose a significant risk in industries such as construction and manufacturing, making helmet use essential for safety. This paper presents a comparative analysis of six YOLO (You Only Look Once) models-YOLOv5mu, YOLOv6m, YOLOv7, YOLOv8m, YOLOv9 (m, GELAN-c, GELAN-e), and YOLOv10m-for helmet detection using a public dataset. The models are evaluated for mean Average Precision (mAP), frames per second (FPS), training time, and model size. YOLOv9 (GELAN-e) achieves the highest accuracy with 96.17% mAP at 0.1% confidence and 82.71% at 70% confidence. YOLOv8m is the fastest, achieving 45.37 FPS, while YOLOv10m has the highest mAP (85.10%) at 70% confidence and is the most compact (33.5 MB). YOLOv9 (GELAN-c) offers the shortest training time (1.32 hours), while YOLOv9 strikes the best balance between accuracy, speed, and size. These findings suggest YOLOv9 as an excellent choice for applications requiring both accuracy and efficiency, while YOLOv8m is optimal for real-time scenarios.

Large-scale biomedical entity recognition and relation extraction are essential foundational tasks for downstream text mining tasks and applications, such as knowledge graph construction. Because many relations span sentence boundaries, document-level entity recognition and relation extraction is closely aligned with real-world demands. However, identifying complex and diverse entities and relations within a limited timeframe is challenging. Therefore, we propose an end-to-end approach called BioECR for biomedical document-level named entity recognition, coreference resolution, and relation extraction. This approach utilizes a bi-encoder structure combined with biomedical entity types and descriptions to solve nested biomedical entities in linear time, thereby enhancing the ability to recognise complex entities and relations. Then a composition graph convolutional neural network was proposed to address the noise in conventional graph convolutional networks, thereby reducing time overhead and selectively fusing multiple entities or contextual information. Finally, by combining entity type clustering methods, the problem of coreference errors among multiple types of entities is solved easily and quickly. Experimental results demonstrate that our approach achieves stateof-the-art performance on all subtasks across three biomedical document-level datasets called CDR, GDA, and BioRED, and our approach reduces the inference time by approximately 60%.

Rank metrics quantify machine learning model prediction errors in the setting of binary classification with a given dataset, but imbalances in the class distribution or model predictions can significantly influence the interpretation. We analyze and interpret the area under the receiver operating characteristic curve (auROC) and an area under the precision recall curve, or average precision score (AP), in the presence of two kinds of imbalance: dataset class imbalance and model prediction imbalance. We show that the auROC can have a symmetric but not necessarily uni-modal response to dataset imbalance, but it is insensitive to model prediction imbalance. The AP has an asymmetric response that we show is also not uni-modal. The AP and auROC scores can be challenging to interpret due to these properties. We propose a novel precision recall space, the PR*, and associated average precision score, called AP*, to address the identified shortcomings of the AP and auROC. The proposed AP* metric has a symmetric response to class imbalance, like the auROC, a sensitivity to model prediction imbalance, like the AP, and it adds both a symmetry to model prediction imbalance and unimodal response to class imbalance that makes it more reliable for model selection than the AP or auROC. For model selection, the AP* gives a higher score to the model that tends to make most of its errors in the majority class.

The simultaneous progress of integrated optical frequency comb (OFC) and radio frequency (RF) photonic signal processing technique have promoted the rapid development of real-time signal processing. Integrated optical frequency comb offer multiple wavelengths as a powerful source for RF photonic signal transversal filter. Here, we review development of real-time signal processing system consisting of integrated OFC and RF photonic signal transversal filter in chronological order, and focus on the applications of this system such as differentiator, integrator, Hilbert transformer, and image processor. We also discuss and present our outlook on more parallel functions and further integration of real-time signal processing system.

This paper builds on previous research on the Single-Turn Crescendo Attack (STCA) by introducing new findings on the effectiveness of a four-turn STCA (STCA-4) against Open AI GPT o1 series (Strawberry) models. Building upon our original work with the STCA-3, we demonstrate that GPT o1 Mini and GPT o1 Preview models are vulnerable to an STCA-4 attack, often generating disallowed content such as hate speech and slurs. In contrast, models like Gemini 1.5 Flash and LLaMA 3 8B showed robust resistance to the same attack. We hypothesize that the success of the STCA-4 on GPT models may be due to the familiarity effect, as the attack leverages language generated by the GPT o1 models themselves, or the slow-thinking ability of the o1 series models. These results highlight the need for improved safeguards in large language models to prevent adversarial exploits that condense multi-turn escalations into a single prompt.

This paper presents a structured framework for real-time quantification of uncertainty measures across di!erent estimation topologies (paths) to ensure reliable estimation under fault conditions. Our multi-stage approach first analyzes multiple redundant estimation pathways to e!ectively isolate fault sources and then reconfigures to the most reliable path based on the theoretically quantified uncertainty measure. Considering all sensor configurations and independent pathways for estimating vehicle states, the resulting structure forms a directed acyclic graph, termed an estimation graph. A reconfigurable estimation scheme is proposed to enhance reliability across diverse fault conditions. The framework leverages a detailed structural analysis of the estimation graph to enhance fault detectability, as shown by detecting the fault by measuring vertical acceleration. By theoretically quantifying fault propagation along each estimation path, the framework enables the real-time selection of the optimal path. The proposed theoretical derivations provide a unified approach to quantifying the e!ects of common soft faults, e.g., bias and excessive noise, by appropriately adapting the influence matrices. Validation using a high-fidelity CarSim model and Monte Carlo simulations confirms the accuracy of the theoretical derivations and demonstrates the framework’s e!ectiveness in localizing fault sources and ensuring reliable estimation under various fault conditions.

Large language models (LLMs) have shown great promise in task execution, but traditional closed-loop systems limit their planning efficiency. Addressing this challenge, we introduce a novel open-loop task planning approach that utilizes historical data and empirical heuristics. This method facilitates informed decision-making during task execution. By incorporating a dynamic feedback mechanism, it enables real-time adjustments based on performance metrics and task progression. Additionally, our framework employs multi-task learning to optimize planning across diverse tasks, allowing prioritization based on contextual relevance and urgency. We have established a robust evaluation metric to measure task completion rates and efficiency improvements. Experimental findings reveal that this method significantly reduces planning overhead while enhancing task execution quality, highlighting the benefits of openloop planning for LLMs in dynamic environments.