Search

A significant advancement in artificial intelligence is the development of large language models (LLMs). Despite opposition and explicit bans by some authorities, LLMs continue to play a transformative role, particularly in education, by improving language understanding and generation capabilities. This study explores LLMs' types, history, and training processes, alongside their application in education, including digital and higher education settings. A novel theoretical framework is proposed to guide the integration of LLMs into education, addressing key challenges such as personalization, ethical concerns, and adaptability. Furthermore, the study presents practical case studies and solutions to barriers, such as data privacy and bias, offering insights into their role in enhancing the teaching–learning process. By providing a systematic analysis and proposing a structured framework, this study advances current knowledge and highlights the significant potential of LLMs in revolutionizing education. [ABSTRACT FROM AUTHOR]

This study presents a learning‐based iterative model predictive control (MPC) scheme for unknown (Lipschitz continuous) nonlinear dynamical systems. The proposed method begins by learning the unknown part of the controlled system using a Gaussian process (GP), which helps derive multi‐step reachable sets that are guaranteed to encompass the actual system states. At each time step in each iteration, the MPC controller calculates a sequence of control inputs that robustly satisfy state and control constraints, as well as terminal constraints based on the GP‐based reachable sets. Then only the first control input is applied to the system. After the iteration, the initial state is reset, and the same procedure is executed with the MPC optimization problem defined by the updated terminal set and cost. As iteration goes on, improvement of the control performance is expected since more data is obtained and the environment is progressively explored. The proposed method provides properties such as recursive feasibility and input to state stability of the goal region under certain assumptions. Moreover, bound on the performance cost in each iteration associated with the implementation of the proposed MPC scheme is also analyzed. The results of the simulation study show that the proposed control scheme can iteratively improve the control performance. [ABSTRACT FROM AUTHOR]

This article proposes an online universal self‐learning control (USLC) algorithm based on a physical performance policy‐optimization neural network, which aims to solve the problem of universal self‐learning optimal control laws for nonlinear systems with various uncertain dynamics. As a key system characterization, this algorithm predicts the discrepancy between the optimal and current control laws by evaluating overall performance in each iterative learning cycle, leveraging an offline‐trained universal policy network. This approach is universal, as it does not rely on an exact system model and can adaptively control performance preferences across various tasks by customizing the physical performance cost weights. Using the established control law‐performance surface and contraction Lyapunov function, the necessary assumptions and proofs for the stable convergence of the system within a three‐dimensional manifold space are provided. To demonstrate the universality of USLC, simulation experiments are conducted on two different systems: a low‐order circuit system and a high‐order variable‐span aircraft attitude control system. The stable control achieved under varying initial values and boundary conditions in each system illustrates the effectiveness of the proposed method. Finally, the limitations of this study are discussed. [ABSTRACT FROM AUTHOR]

Abstract Artificial intelligence-driven Chatbots, especially large language models (LLMs) like GPT-4, represent significant progress in digital education. These models excel in mimicking human-like text and transforming learning and teaching methods. This study examines the development, application, and impact of LLMs in education. It highlights their role in automating instructional tasks and promoting personalized learning experiences. Despite integration concerns and ethical debates, LLMs showcase the potential of AI to improve educational practices. Our research concludes that LLMs offer transformative opportunities for education. However, their incorporation requires careful ethical considerations, data privacy measures, and a balance between human educators and AI technologies. The findings suggest strategies for integrating LLMs into educational frameworks to enhance learning outcomes while preserving educational integrity.

Abstract A significant advancement in artificial intelligence is the development of large language models (LLMs). Despite opposition and explicit bans by some authorities, LLMs continue to play a transformative role, particularly in education, by improving language understanding and generation capabilities. This study explores LLMs’ types, history, and training processes, alongside their application in education, including digital and higher education settings. A novel theoretical framework is proposed to guide the integration of LLMs into education, addressing key challenges such as personalization, ethical concerns, and adaptability. Furthermore, the study presents practical case studies and solutions to barriers, such as data privacy and bias, offering insights into their role in enhancing the teaching–learning process. By providing a systematic analysis and proposing a structured framework, this study advances current knowledge and highlights the significant potential of LLMs in revolutionizing education.

Abstract This article studies cooperative adaptive cruise control (CACC) for vehicle platoons with consideration of the unknown nonlinear vehicle dynamics that are normally ignored in the literature. A unified data‐driven CACC design is proposed for platoons of pure automated vehicles (AVs) or of mixed AVs and human‐driven vehicles (HVs). The CACC leverages online‐collected sufficient data samples of vehicle accelerations, spacing, and relative velocities. The data‐driven control design is formulated as a semidefinite program that can be solved efficiently using off‐the‐shelf solvers. Efficacy of the proposed CACC are demonstrated on a platoon of pure AVs and mixed platoons with different penetration rates of HVs using a representative aggressive driving profile. Advantage of the proposed design is also shown through a comparison with the classic adaptive cruise control (ACC) method.

Abstract The integration of renewable energy sources (RESs) into power systems has increased significantly due to technical, economic, and environmental factors, necessitating greater flexibility to manage variable consumption loads and renewable energy generation. Accurate forecasting of solar energy production and consumption load is critical for enhancing power system flexibility. This study introduces a novel deep learning model, a spatial‐temporal hybrid convolutional‐transformer (CT‐Transformer) network with unique features and extended memory capacity. Additionally, a flexibility index (FI) is introduced to evaluate power system flexibility (PSF) based on the forecasting results. The CT‐Transformer forecasts PSF for the next 24 and 168 hours, using the FI to evaluate PSF based on forecasting results. The input data includes meteorological, solar energy production, load demand, and pricing data from France, comprising hourly data from 2015 and 2016 (about 17,520 entries). Data preprocessing involves correcting incomplete and irrelevant segments. The CT‐Transformer's performance is compared to other deep learning techniques, showing superior results with the lowest prediction error (2.5%) and a maximum error of 10.1% (MAE). It also achieved a prediction error of 0.08% for system flexibility, compared to the highest error of 0.96%. This research highlights the CT‐Transformer's potential for accurate RES and load forecasting and PSF evaluation.

Abstract The complexity of neuronal networks, characterized by interconnected neurons, presents significant challenges in control due to their nonlinear and intricate behaviour. This paper introduces a novel method designed to generate control inputs for neuronal networks to regulate the firing patterns of modules within the network. This methodology is built upon temporal deep unfolding‐based model predictive control, a technique rooted in the deep unfolding method commonly used in wireless signal processing. To address the unique dynamics of neurons, such as zero gradients in firing times, the method employs approximations of input currents using a sigmoid function during its development. The effectiveness of this approach is validated through extensive numerical simulations. Furthermore, control experiments were conducted by reducing the number of input neurons to identify critical features for control. Various selection techniques were utilized to pinpoint key input neurons. These experiments shed light on the importance of specific input neurons in controlling module firing within neuronal networks. Thus, this study presents a tailored methodology for managing networked neurons, extends temporal deep unfolding‐based model predictive control to nonlinear systems with reset dynamics, and demonstrates its ability to achieve desired firing patterns in neuronal networks.

Abstract Trained deep reinforcement learning (DRL) based controllers can effectively control dynamic systems where classical controllers can be ineffective and difficult to tune. However, the lack of closed‐loop stability guarantees of systems controlled by trained DRL agents hinders their adoption in practical applications. This research study investigates the closed‐loop stability of dynamic systems controlled by trained DRL agents using Lyapunov analysis based on a linear‐quadratic polynomial approximation of the trained agent. In addition, this work develops an understanding of the system's stability margin to determine operational boundaries and critical thresholds of the system's physical parameters for effective operation. The proposed analysis is verified on a DRL‐controlled system for several simulated and experimental scenarios. The DRL agent is trained using a detailed dynamic model of a non‐linear system and then tested on the corresponding real‐world hardware platform without any fine‐tuning. Experiments are conducted on a wide range of system states and physical parameters and the results have confirmed the validity of the proposed stability analysis (https://youtu.be/QlpeD5sTlPU).

This article studies cooperative adaptive cruise control (CACC) for vehicle platoons with consideration of the unknown nonlinear vehicle dynamics that are normally ignored in the literature. A unified data‐driven CACC design is proposed for platoons of pure automated vehicles (AVs) or of mixed AVs and human‐driven vehicles (HVs). The CACC leverages online‐collected sufficient data samples of vehicle accelerations, spacing, and relative velocities. The data‐driven control design is formulated as a semidefinite program that can be solved efficiently using off‐the‐shelf solvers. Efficacy of the proposed CACC are demonstrated on a platoon of pure AVs and mixed platoons with different penetration rates of HVs using a representative aggressive driving profile. Advantage of the proposed design is also shown through a comparison with the classic adaptive cruise control (ACC) method. [ABSTRACT FROM AUTHOR]

The integration of renewable energy sources (RESs) into power systems has increased significantly due to technical, economic, and environmental factors, necessitating greater flexibility to manage variable consumption loads and renewable energy generation. Accurate forecasting of solar energy production and consumption load is critical for enhancing power system flexibility. This study introduces a novel deep learning model, a spatial‐temporal hybrid convolutional‐transformer (CT‐Transformer) network with unique features and extended memory capacity. Additionally, a flexibility index (FI) is introduced to evaluate power system flexibility (PSF) based on the forecasting results. The CT‐Transformer forecasts PSF for the next 24 and 168 hours, using the FI to evaluate PSF based on forecasting results. The input data includes meteorological, solar energy production, load demand, and pricing data from France, comprising hourly data from 2015 and 2016 (about 17,520 entries). Data preprocessing involves correcting incomplete and irrelevant segments. The CT‐Transformer's performance is compared to other deep learning techniques, showing superior results with the lowest prediction error (2.5%) and a maximum error of 10.1% (MAE). It also achieved a prediction error of 0.08% for system flexibility, compared to the highest error of 0.96%. This research highlights the CT‐Transformer's potential for accurate RES and load forecasting and PSF evaluation. [ABSTRACT FROM AUTHOR]

To address the variable initial state and trail length this paper first presents a robust PD‐type open‐closed‐loop iterative learning control (ILC) law for a multiple‐input‐multiple‐output (MIMO) linear discrete‐time system. It is demonstrated that the convergence condition is dependent on the PD‐type feed‐forward learning gains, while an appropriate feedback learning gain can improve the ILC convergence performance. As a special case of PD‐type open‐closed‐loop ILC law, P‐type and D‐type open‐closed‐loop ILC laws are deduced. The three developed ILC laws ensure that as the number of iterations approaches infinity, the expectation of ILC tracking error will be constrained within a limited range, where the boundary is proportional to the initial state variation. Through a numerical simulation, the effectiveness of the proposed ILC laws is illustrated. [ABSTRACT FROM AUTHOR]

The complexity of neuronal networks, characterized by interconnected neurons, presents significant challenges in control due to their nonlinear and intricate behaviour. This paper introduces a novel method designed to generate control inputs for neuronal networks to regulate the firing patterns of modules within the network. This methodology is built upon temporal deep unfolding‐based model predictive control, a technique rooted in the deep unfolding method commonly used in wireless signal processing. To address the unique dynamics of neurons, such as zero gradients in firing times, the method employs approximations of input currents using a sigmoid function during its development. The effectiveness of this approach is validated through extensive numerical simulations. Furthermore, control experiments were conducted by reducing the number of input neurons to identify critical features for control. Various selection techniques were utilized to pinpoint key input neurons. These experiments shed light on the importance of specific input neurons in controlling module firing within neuronal networks. Thus, this study presents a tailored methodology for managing networked neurons, extends temporal deep unfolding‐based model predictive control to nonlinear systems with reset dynamics, and demonstrates its ability to achieve desired firing patterns in neuronal networks. [ABSTRACT FROM AUTHOR]

Trained deep reinforcement learning (DRL) based controllers can effectively control dynamic systems where classical controllers can be ineffective and difficult to tune. However, the lack of closed‐loop stability guarantees of systems controlled by trained DRL agents hinders their adoption in practical applications. This research study investigates the closed‐loop stability of dynamic systems controlled by trained DRL agents using Lyapunov analysis based on a linear‐quadratic polynomial approximation of the trained agent. In addition, this work develops an understanding of the system's stability margin to determine operational boundaries and critical thresholds of the system's physical parameters for effective operation. The proposed analysis is verified on a DRL‐controlled system for several simulated and experimental scenarios. The DRL agent is trained using a detailed dynamic model of a non‐linear system and then tested on the corresponding real‐world hardware platform without any fine‐tuning. Experiments are conducted on a wide range of system states and physical parameters and the results have confirmed the validity of the proposed stability analysis (https://youtu.be/QlpeD5sTlPU). [ABSTRACT FROM AUTHOR]

Recommender systems -RS- have experienced exponential growth in various fields, especially in the tourism sector, improving tourism activities’ accuracy, personalization, and experience, thus strengthening indicators such as promotion. However, some challenges and opportunities exist to overcome, such as the lack of data on emerging destinations wishing to adopt these solutions. This manuscript presents a literature review of the current trends in RS applied to the tourism industry, including categories associated with their use and emerging techniques. Likewise, it presents a pathway for implementing an RS when insufficient data are available for a destination. The SLR followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and used the WoS, Science Direct, and Scopus databases. The results show that the hybrid RS integrates deep learning algorithms, data analytics, and optimisation techniques with collaborative tourism features to provide innovative solutions in terms of performance, accuracy, and personalisation of recommendations, thus achieving the management of tourist destinations or tourism-oriented services. Emerging destinations that lack RS data in tourism should use various data sources generated by tourists on social media, tourism portals, and through their interaction with tour operators. New tourism recommender system solutions can emerge following trends integrating new technologies based on user experience, collaboration, and the integration of multiple data sources.

This letter addresses the output consensus problem for a class of two‐dimensional (2D) multi agent systems (MASs) described by Fornasini–Marchesini model. By transforming the 2D agent into a compact form, an adaptive variable, which adjusted by the tracking errors of itself and the neighbour agents, is designed to approximate the unknown varying coefficient. Then, based on the approximated coefficient and the iteration‐varying reference surfaces, the distributed adaptive iterative learning control strategy is obtained. The output consensus of the 2D multi agent is proved. Simulations are included to verify the effectiveness of the investigated distributed adaptive ILC for 2D MASs with random variations on initial condition and reference surface. [ABSTRACT FROM AUTHOR]

Abstract This paper presents the use of one-way analysis of variance ANOVA as an effective tool for ranking the features calculated from diagnostic signals and evaluates their impact on the accuracy of the machine learning system's classification of displacement pump wear.The first part includes a review of contemporary diagnostic systems and a description of typical damage of multi-piston displacement pumps and Its causes. The work also contains description of a diagnostic experiment which was conducted in order to obtain the matrix of vibration signals and the matrix of pressures measured at selected locations on the pump housing and at the pump pressure line. The measured signals were subjected to time–frequency analysis. The features of signals calculated in the time and frequency domains were ranked using the ANOVA. The next step involved the use the available classifiers in pump wear evaluation, conducting tests and assessing their effectiveness in terms of the ranking of features and the origin of diagnostic signals.

Advancements in information and communications technology (ICT) have fundamentally transformed computing, notably through the internet of things (IoT) and its healthcare-focused branch, the internet of medical things (IoMT). These technologies, while enhancing daily life, face significant security risks, including ransomware. To counter this, the authors present a scalable, hybrid machine learning framework that effectively identifies IoMT ransomware attacks, conserving the limited resources of IoMT devices. To assess the effectiveness of their proposed solution, the authors undertook an experiment using a state-of-the-art dataset. Their framework demonstrated superiority over conventional detection methods, achieving an impressive 87% accuracy rate. Building on this foundation, the framework integrates a multi-faceted feature extraction process that discerns between benign and malign actions, with a subsequent in-depth analysis via a neural network. This advanced analysis is pivotal in precisely detecting and terminating ransomware threats, offering a robust solution to secure the IoMT ecosystem. [ABSTRACT FROM AUTHOR]

This research paper intends to provide real-life applications of Generative AI (GAI) in the cybersecurity domain. The frequency, sophistication and impact of cyber threats have continued to rise in today’s world. This ever-evolving threat landscape poses challenges for organizations and security professionals who continue looking for better solutions to tackle these threats. GAI technology provides an effective way for them to address these issues in an automated manner with increasing efficiency. It enables them to work on more critical security aspects which require human intervention, while GAI systems deal with general threat situations. Further, GAI systems can better detect novel malware and threatening situations than humans. This feature of GAI, when leveraged, can lead to higher robustness of the security system. Many tech giants like Google, Microsoft etc., are motivated by this idea and are incorporating elements of GAI in their cybersecurity systems to make them more efficient in dealing with ever-evolving threats. Many cybersecurity tools like Google Cloud Security AI Workbench, Microsoft Security Copilot, SentinelOne Purple AI etc., have come into the picture, which leverage GAI to develop more straightforward and robust ways to deal with emerging cybersecurity perils. With the advent of GAI in the cybersecurity domain, one also needs to take into account the limitations and drawbacks that such systems have. This paper also provides some of the limitations of GAI, like periodically giving wrong results, costly training, the potential of GAI being used by malicious actors for illicit activities etc.

The digital competences of university students are developed using a digital learning ecosystem that integrates: 1) virtual learning environments; 2) digital learning tools; and 3) learning methodologies. This research followed the methodology of systematic literature mapping and review, searching the WoS and Scopus databases and obtaining a total of 5,652 articles between 2001 and 2023. Inclusion and exclusion criteria were then applied to reduce the number of selected articles and carry out a systematic literature mapping and review. Among the relevant results of the literature mapping and systematic review, the geographic distribution of scientific publications, the educational areas in which they have been worked on, and the universities were identified. Educational methodologies, technological tools, and virtual learning environments used to develop university students’ digital competences holistically were also determined. This study is useful as it provides a comprehensive, general, and detailed overview of scientific production and its main contributions regarding the methodologies, tools, and environments that contribute to developing students’ the digital competences in higher education.

Abstract Concerning the problems that the traditional Convolutional Neural Network (CNN) ignores contextual semantic information, and the traditional Recurrent Neural Network (RNN) has information memory loss and vanishing gradient, this paper proposes a Bi‐directional Encoder Representations from Transformers (BERT)‐based dual‐channel parallel hybrid neural network model for text sentiment analysis. The BERT model is used to convert text into word vectors; the dual‐channel parallel hybrid neural network model constructed by CNN and Bi‐directional Long Short‐Term Memory (BiLSTM) extracts local and global semantic features of the text, which can obtain more comprehensive sentiment features; the attention mechanism enables some words to get more attention that highlights important words and improves the model's sentiment classification ability. Finally, the dual‐channel output features are fused for sentiment classification. The experimental results on the hotel review datasets show that the Accuracy of the proposed model in sentiment classification reaches 92.35% and the F1 score reaches 91.59%.

Introduction: The COVID‐19 pandemic revealed numerous barriers to effectively managing public health crises, including difficulties in using publicly available, community‐level data to create learning systems in support of local public health decision responses. Early in the COVID‐19 pandemic, a group of health care partners began meeting to learn from their collective experiences. We identified key tools and processes for using data and learning system structures to drive equitable public health decision making throughout different phases of the pandemic. Methods: In fall of 2021, the team developed an initial theory of change directed at achieving herd immunity for COVID‐19. The theoretical drivers were explored qualitatively through a series of nine 45‐min telephonic interviews conducted with 16 public health and community leaders across the United States. Interview responses were analyzed into key themes to inform potential future practices, tools, and systems. In addition to the interviews, partners in Dallas and Cincinnati reflected on their own COVID‐19 experiences. Results: Interview responses fell broadly into four themes that contribute to effective, community driven responses to COVID‐19: real‐time, accessible data that are mindful of the tension between community transparency and individual privacy; a continued fostering of public trust; adaptable infrastructures and systems; and creating cohesive community coalitions with shared alignment and goals. These themes and partner experiences helped us revise our preliminary theory of change around the importance of community collaboration and trust building and also helped refine the development of the Community Protection Dashboard tool. Conclusions: There was broad agreement amongst public health and community leaders about the key elements of the data and learning systems required to manage public health responses to COVID‐19. These findings may be informative for guiding the use of data and learning in the management of future public health crises or population health initiatives. [ABSTRACT FROM AUTHOR]

Background The presence of artificial intelligence (AI) in many areas of social life is becoming widespread. The advantages of AI are being observed in medicine, commerce, automobiles, customer service, agriculture and production in factory settings, among others. Workers first encountered robots in the work environment in the 1960s. Since then, intelligent systems have become much more advanced. The expansion of AI functionality in the work environment exacerbates human health risks. These can be physical (lack of adequate machine control, accidents) or psychological (technostress, fear, automation leading to job exclusion, changes in the labour market, widening social differences). Material and Methods The purpose of this article is to identify, based on selected literature, possible applications of AI and the potential benefits and risks for humans. Results The main area of interest was the contemporary work environment and the health consequences associated with access to smart technologies. A key research area for us was the relationship between AI and increased worker control. Conclusions In the article, the authors emphasize the importance of relevant EU legislation that guarantees respect for the rights of the employed. The authors put forward the thesis that the new reality with the widespread use of AI, requires an analysis of its impact on the human psycho-social and health situation. Thus, a legal framework defining the scope of monitoring and collection of sensitive data is necessary. Med Pr. 2023;74(3):227–33

The relevance of the declared subject of this research work is determined by the need to develop and implement software for users of various skill levels, as well as to create effective multi-scenario systems for describing processes occurring in multiple environments. The primary purpose of this scientific research is to study the principles of modeling the functioning of multi-scenario systems. The basis of the methodological approach in this scientific study is a combination of methods of system analysis of the principles of creating models for the functioning of multi-scenario systems with an analytical study of the prospects for building monolithic architectures of online learning systems. In the course of this scientific research, results were obtained that describe the principles of modeling a multi-scenario online learning system, as well as illustrating the features of the interaction of individual subsystems within a single multi-scenario system, taking into account the effectiveness of each of the subsystems performing the functions assigned to it within a single multi-scenario system. The results obtained reflect the fundamental principles of building the operation of multi-scenario systems in the conditions of the need to process a large amount of data, taking into account the difference in user characteristics, their levels of preparedness, as well as the variety of user requests that have a significant impact on the process of creating a multi-scenario system model and its functioning in constantly changing environments. External conditions. The practical significance of the results obtained in this scientific study, as well as the conclusions formulated on their basis, lies in the possibility of their application in the development of information presentation systems, the operation of which is based on the principle of multi-scenario, in order to provide the option of choosing modes of use and their automatic adjustment.

Ahmad H Alwashmi Department of Orthopedic Surgery, College of Medicine, Qassim University, Buraydah, 52571, Saudi ArabiaCorrespondence: Ahmad H Alwashmi, Department of Orthopedic Surgery, College of Medicine, Qassim University, Buraydah, Qassim, 52571, Saudi Arabia, Email a.alwashmi@qu.edu.saBackground: Residents in training must employ a variety of study strategies, as they not only participate in academic studies but also interact with patients. This study aimed to evaluate the study practices and factors affecting those practices among Saudi Arabian physical medicine and rehabilitation residents during their residency program.Methods: In this cross-sectional study, a previously used questionnaire was distributed to Saudi Arabian physiatry residents from July 1 to August 15, 2022, via a social media platform and completed using a Google Forms survey. A Microsoft Excel spreadsheet was used to collect, clean, and import the data before IBM SPSS Statistics for Windows, version 22.0 was utilized for statistical analysis.Results: The data of 94.91% of respondents were included in the analysis. Individuals who were female, unmarried or divorced, and without children predominated. Only 17.9% (n = 10) of the residents believed that their training program effectively prepared them to pass the board examination, which was the most strongly motivating factor for studying for 85.7% of respondents. Over two-thirds of the residents mentioned that they regularly exercise. Residents who studied more than 11 hours per week had a significantly lower score in the category of factors that negatively affect examination performance (M = 12.33 ± 2.82, F = 2.794, P < 0.05). Females, final-year residents, and Riyadh residents studied more than their counterparts.Conclusion: Our study is the first to investigate how Saudi physiatrists study, with the finding that current physiatry residents employ a combination of traditional and contemporary learning strategies. This information can help stakeholders to understand current training challenges, improve the quality of training for physiatry residents, and create an ideal learning environment.Keywords: cross-sectional study, questionnaire, performance assessment, students, learning systems, physiatry

Concerning the problems that the traditional Convolutional Neural Network (CNN) ignores contextual semantic information, and the traditional Recurrent Neural Network (RNN) has information memory loss and vanishing gradient, this paper proposes a Bi‐directional Encoder Representations from Transformers (BERT)‐based dual‐channel parallel hybrid neural network model for text sentiment analysis. The BERT model is used to convert text into word vectors; the dual‐channel parallel hybrid neural network model constructed by CNN and Bi‐directional Long Short‐Term Memory (BiLSTM) extracts local and global semantic features of the text, which can obtain more comprehensive sentiment features; the attention mechanism enables some words to get more attention that highlights important words and improves the model's sentiment classification ability. Finally, the dual‐channel output features are fused for sentiment classification. The experimental results on the hotel review datasets show that the Accuracy of the proposed model in sentiment classification reaches 92.35% and the F1 score reaches 91.59%. [ABSTRACT FROM AUTHOR]

This article considers the subject of information losses arising from the finite data sets used in the training of neural classifiers. It proves a relationship between such losses as the product of the expected total variation of the estimated neural model with the information about the feature space contained in the hidden representation of that model. It then bounds this expected total variation as a function of the size of randomly sampled data sets in a fairly general setting, and without bringing in any additional dependence on model complexity. It ultimately obtains bounds on information losses that are less sensitive to input compression and in general much smaller than existing bounds. This article then uses these bounds to explain some recent experimental findings of information compression in neural networks that cannot be explained by previous work. Finally, this article shows that not only are these bounds much smaller than existing ones, but they also correspond well with experiments.

Learning Analytics (LA) is increasingly used in Education to set prediction models from artificial intelligence to determine learning profiles. This study aims to determine to what extent K-nearest neighbor and random forest algorithms could become a useful tool for improving the teaching-learning process and reducing academic failure in two Physics courses at the Technological Institute of Monterrey, México (n = 268). A quasi-experimental and mixed method approach was conducted. The main results showed significant differences between the first and second term evaluations in the two groups. One of the main findings of the study is that the predictions were not very accurate for each student in the first term evaluation. However, the predictions became more accurate as the algorithm was fed with larger datasets from the second term evaluation. This result indicates how predictive algorithms based on decision trees, can offer a close approximation to the academic performance that will occur in the class, and this information could be use along with the personal impressions coming from the teacher.

In modern networking research, infrastructure-assisted unmanned autonomous vehicles (UAVs) are actively considered for real-time learning-based surveillance and aerial data-delivery under unexpected 3D free mobility and coordination. In this system model, it is essential to consider the power limitation in UAVs and autonomous object recognition (for abnormal behavior detection) deep learning performance in infrastructure/towers. To overcome the power limitation of UAVs, this paper proposes a novel aerial scheduling algorithm between multi-UAVs and multi-towers where the towers conduct wireless power transfer toward UAVs. In addition, to take care of the high-performance learning model training in towers, we also propose a data delivery scheme which makes UAVs deliver the training data to the towers fairly to prevent problems due to data imbalance (e.g., huge computation overhead caused by larger data delivery or overfitting from less data delivery). Therefore, this paper proposes a novel workload-aware scheduling algorithm between multi-towers and multi-UAVs for joint power-charging from towers to their associated UAVs and training data delivery from UAVs to their associated towers. To compute the workload-aware optimal scheduling decisions in each unit time, our solution approach for the given scheduling problem is designed based on Markov decision process (MDP) to deal with (i) time-varying low-complexity computation and (ii) pseudo-polynomial optimality. As shown in performance evaluation results, our proposed algorithm ensures (i) sufficient times for resource exchanges between towers and UAVs, (ii) the most even and uniform data collection during the processes compared to the other algorithms, and (iii) the performance of all towers convergence to optimal levels.

AI and Machine Learning (ML) offer powerful tools to support clinical decision making in emergency situations such as the COVID-19 pandemic. In this context, the application of ML requires to design predictive systems that have adequate accuracy and can effectively deal with issues concerning data quality, sensitive errors, uncertainty, and interpretability of the predictions. We present a methodology that deals with all these problems and a concrete study of its application to estimate the prognosis of hospitalised patients with COVID-19. In particular, we address the task of predicting the outcome (alive or deceased) of a patient at different times of her/his hospitalisation minimising false negatives (wrong survival predictions). The proposed methodology builds different optimised ML models to select those that perform the best to recognise, at different times of hospitalisation, patients who will have an unfavourable prognosis (decease). These models exploit a new algorithm, presented in the paper, that identifies an uncertainty threshold to rule out uncertain predictions with the purpose of making a ML model both more performing and more reliable. Moreover, we propose a general method for automatically extracting multi-variable prognostic rules from the available data. Such rules can provide possible new useful knowledge on the considered disease. We also show how they can be used effectively to explain the predictions made by the ML models. All proposed methods and techniques are experimentally evaluated in the context of our application task.

When adopting a model-based formulation, solving inverse problems encountered in multiband imaging requires to define spatial and spectral regularizations. In most of the works of the literature, spectral information is extracted from the observations directly to derive data-driven spectral priors. Conversely, the choice of the spatial regularization often boils down to the use of conventional penalizations (e.g., total variation) promoting expected features of the reconstructed image (e.g., piece-wise constant). In this work, we propose a generic framework able to capitalize on an auxiliary acquisition of high spatial resolution to derive tailored data-driven spatial regularizations. This approach leverages on the ability of deep learning to extract high level features. More precisely, the regularization is conceived as a deep generative network able to encode spatial semantic features contained in this auxiliary image of high spatial resolution. To illustrate the versatility of this approach, it is instantiated to conduct two particular tasks, namely multiband image fusion and multiband image inpainting. Experimental results obtained on these two tasks demonstrate the benefit of this class of informed regularizations when compared to more conventional ones. [ABSTRACT FROM AUTHOR]

Recently, feature relation learning has attracted extensive attention in cross-spectral image patch matching. However, most feature relation learning methods can only extract shallow feature relations and are accompanied by the loss of useful discriminative features or the introduction of disturbing features. Although the latest multi-branch feature difference learning network can relatively sufficiently extract useful discriminative features, the multi-branch network structure it adopts has a large number of parameters. Therefore, we propose a novel two-branch feature interaction learning network (FIL-Net). Specifically, a novel feature interaction learning idea for cross-spectral image patch matching is proposed, and a new feature interaction learning module is constructed, which can effectively mine common and private features between cross-spectral image patches, and extract richer and deeper feature relations with invariance and discriminability. At the same time, we re-explore the feature extraction network for the cross-spectral image patch matching task, and a new two-branch residual feature extraction network with stronger feature extraction capabilities is constructed. In addition, we propose a new multi-loss strong-constrained optimization strategy, which can facilitate reasonable network optimization and efficient extraction of invariant and discriminative features. Furthermore, a public VIS-LWIR patch dataset and a public SEN1-2 patch dataset are constructed. At the same time, the corresponding experimental benchmarks are established, which are convenient for future research while solving few existing cross-spectral image patch matching datasets. Extensive experiments show that the proposed FIL-Net achieves state-of-the-art performance in three different cross-spectral image patch matching scenarios. [ABSTRACT FROM AUTHOR]

Near-infrared and visible face recognition (NIR-VIS) is attracting increasing attention because of the need to achieve face recognition in low-light conditions to enable 24-hour secure retrieval. However, annotating identity labels for a large number of heterogeneous face images is time-consuming and expensive, which limits the application of the NIR-VIS face recognition system to larger scale real-world scenarios. In this paper, we attempt to achieve NIR-VIS face recognition in an unsupervised domain adaptation manner. To get rid of the reliance on manual annotations, we propose a novel Robust cross-domain Pseudo-labeling and Contrastive learning (RPC) network which consists of three key components, i.e., NIR cluster-based Pseudo labels Sharing (NPS), Domain-specific cluster Contrastive Learning (DCL) and Inter-domain cluster Contrastive Learning (ICL). Firstly, NPS is presented to generate pseudo labels by exploring robust NIR clusters and sharing reliable label knowledge with VIS domain. Secondly, DCL is designed to learn intra-domain compact yet discriminative representations. Finally, ICL dynamically combines and refines intrinsic identity relationships to guide the instance-level features to learn robust and domain-independent representations. Extensive experiments are conducted to verify an accuracy of over 99% in pseudo label assignment and the advanced performance of RPC network on four mainstream NIR-VIS datasets. [ABSTRACT FROM AUTHOR]

In image processing, images are usually composed of partial views due to the uncertainty of collection and how to efficiently process these images, which is called incomplete multi-view learning, has attracted widespread attention. The incompleteness and diversity of multi-view data enlarges the difficulty of annotation, resulting in the divergence of label distribution between the training and testing data, named as label shift. However, existing incomplete multi-view methods generally assume that the label distribution is consistent and rarely consider the label shift scenario. To address this new but important challenge, we propose a novel framework termed as Incomplete Multi-view Learning under Label Shift (IMLLS). In this framework, we first give the formal definitions of IMLLS and the bidirectional complete representation which describes the intrinsic and common structure. Then, a multilayer perceptron which combines the reconstruction and classification loss is employed to learn the latent representation, whose existence, consistency and universality are proved with the theoretical satisfaction of label shift assumption. After that, to align the label distribution, the learned representation and trained source classifier are used to estimate the importance weight by designing a new estimation scheme which balances the error generated by finite samples in theory. Finally, the trained classifier reweighted by the estimated weight is fine-tuned to reduce the gap between the source and target representations. Extensive experimental results validate the effectiveness of our algorithm over existing state-of-the-arts methods in various aspects, together with its effectiveness in discriminating schizophrenic patients from healthy controls. [ABSTRACT FROM AUTHOR]

Matching landmark patches from a real-time image captured by an on-vehicle camera with landmark patches in an image database plays an important role in various computer perception tasks for autonomous driving. Current methods focus on local matching for regions of interest and do not take into account spatial neighborhood relationships among the image patches, which typically correspond to objects in the environment. In this paper, we construct a spatial graph with the graph vertices corresponding to patches and edges capturing the spatial neighborhood information. We propose a joint feature and metric learning model with graph-based learning. We provide a theoretical basis for the graph-based loss by showing that the information distance between the distributions conditioned on matched and unmatched pairs is maximized under our framework. We evaluate our model using several street-scene datasets and demonstrate that our approach achieves state-of-the-art matching results. [ABSTRACT FROM AUTHOR]

In order to adapt to the needs of the times, many enterprises have combined AI (Artificial Intelligence) technology with enterprise management to enhance their competitiveness, which is also essential in the field of education. This article attempted to integrate AI technology into multimedia teaching to improve the teaching quality of teachers and students' understanding of knowledge. The experiment found that AI technology could effectively improve students' understanding of knowledge points and also improve the clarity of knowledge point explanations during the teaching process. It could be seen that using AI intelligent learning systems for multimedia demonstrations could greatly improve the teaching quality of teachers, as well as the learning interest and effectiveness of students. [ABSTRACT FROM AUTHOR]

Copyright of Tarling: Journal of Language Education is the property of UIN Saizu Purwokerto and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The relevance of the declared subject of this research work is determined by the need to develop and implement software for users of various skill levels, as well as to create effective multi-scenario systems for describing processes occurring in multiple environments. The primary purpose of this scientific research is to study the principles of modeling the functioning of multi-scenario systems. The basis of the methodological approach in this scientific study is a combination of methods of system analysis of the principles of creating models for the functioning of multi-scenario systems with an analytical study of the prospects for building monolithic architectures of online learning systems. In the course of this scientific research, results were obtained that describe the principles of modeling a multi-scenario online learning system, as well as illustrating the features of the interaction of individual subsystems within a single multi-scenario system, taking into account the effectiveness of each of the subsystems performing the functions assigned to it within a single multi-scenario system. The results obtained reflect the fundamental principles of building the operation of multi-scenario systems in the conditions of the need to process a large amount of data, taking into account the difference in user characteristics, their levels of preparedness, as well as the variety of user requests that have a significant impact on the process of creating a multi-scenario system model and its functioning in constantly changing environments. External conditions. The practical significance of the results obtained in this scientific study, as well as the conclusions formulated on their basis, lies in the possibility of their application in the development of information presentation systems, the operation of which is based on the principle of multi-scenario, in order to provide the option of choosing modes of use and their automatic adjustment. [ABSTRACT FROM AUTHOR]

The abundance of online educational resources has made it increasingly difficult for students to identify the correct learning materials in recent years. Overcoming the information overload that has emerged in the new education systems is possible via a tailored recommendation system. It encourages students to look for new ways to get around the subject matter and to use information from all across the world. Because of this, many academics are working to create learning systems that incorporate methods for creating a unique learning experience for each user. Therefore, our proposed approach was to create an appropriate learning route for each student, and they are using Educational Resource Information Based on an Adaptable Genetic Algorithm(ERI-AGA). Evidence from studies shows that the suggested technique can provide relevant course materials for students based on the specific needs of students to help them study better in a Web-based system. Personal recommendation engine, pre-processing and learning-based model development, and implementation of the recommendation system will be researched. Participatory budgeting PB-level data storage and processing as well as the ability to suggest in real time will be studied. The capacity to make real-time suggestions and the storing and processing of PB-level data will be investigated. It was critical to check the system's availability by running associated tasks and performance tests. The comparison values demonstrated that ERI-AGA was a reliable and accurate assessment procedure. [ABSTRACT FROM AUTHOR]

Abstract Both global dependency and local correlation are crucial for solving the scale variation of crowd. However, most of previous methods fail to take two factors into consideration simultaneously. Against the aforementioned issue, a deformable channel non‐local network, abbreviated as DCNLNet for crowd counting, which can simultaneously learn global context information and adaptive local receptive field is proposed. Specifically, the proposed DCNLNet consists of two well‐crafted designed modules: deformable channel non‐local block (DCNL) and spatial attention feature fusion block (SAFF). The DCNL encodes long‐range dependencies between pixels and the adaptive local correlation with channel non‐local and deformable convolution, respectively, benefiting for improving the spatial discrimination of features. While the SAFF aims to aggregate the cross‐level information, which interacts these features from different depths and learns specific weights for the feature maps with spatial attention. Extensive experiments are performed on three crowd counting benchmark datasets and experimental results indicate that the proposed DCNLNet achieves compelling performance compared to other representative counting models.

Reinforcement learning (RL) has emerged as an effective approach for building an intelligent system, which involves multiple self-operated agents to collectively accomplish a designated task. More importantly, there has been a renewed focus on RL since the introduction of deep learning that essentially makes RL feasible to operate in high-dimensional environments. However, there are many diversified research directions in the current literature, such as multi-agent and multi-objective learning, and human-machine interactions. Therefore, in this paper, we propose a comprehensive software architecture that not only plays a vital role in designing a connect-the-dots deep RL architecture but also provides a guideline to develop a realistic RL application in a short time span. By inheriting the proposed architecture, software managers can foresee any challenges when designing a deep RL-based system. As a result, they can expedite the design process and actively control every stage of software development, which is especially critical in agile development environments. For this reason, we design a deep RL-based framework that strictly ensures flexibility, robustness, and scalability. To enforce generalization, the proposed architecture also does not depend on a specific RL algorithm, a network configuration, the number of agents, or the type of agents. [ABSTRACT FROM AUTHOR]

Recent legislation reforming the oversight of Oranga Tamariki and the role of the children's commissioner was met with all but universal opposition. A key concern was that locating monitoring of the care and protection of children with a government department (and not the commissioner) was too close to ministers to ensure the level of independence required for such a function. This article suggests that the public sector policy advisory system was not robust enough to come up with the optimal policy solution when, in effect, all others said it was wrong. The case gives cause for the public sector to reflect upon the quality of its advisory function. [ABSTRACT FROM AUTHOR]

Hydraulic power systems are commonly used in heavy industry (usually highly energy-intensive) and are often associated with high power losses. Designing a suitable system to allow an early assessment of the wear conditions of components in a hydraulic system (e.g., an axial piston pump) can effectively contribute to reducing energy losses during use. This paper presents the application of a deep machine learning system to determine the efficiency state of a multi-piston positive displacement pump. Such pumps are significant in high-power hydraulic systems. The correct operation of the entire hydraulic system often depends on its proper functioning. The wear and tear of individual pump components usually leads to a decrease in the pump's operating pressure and volumetric losses, subsequently resulting in a decrease in overall pump efficiency and increases in vibration and pump noise. This in turn leads to an increase in energy losses throughout the hydraulic system, which releases excess heat. Typical failures of the discussed pumps and their causes are described after reviewing current research work using deep machine learning. Next, the test bench on which the diagnostic experiment was conducted and the selected operating signals that were recorded are described. The measured signals were subjected to a time–frequency analysis, and their features, calculated in terms of the time and frequency domains, underwent a significance ranking using the minimum redundancy maximum relevance (MRMR) algorithm. The next step was to design a neural network structure to classify the wear state of the pump and to test and evaluate the effectiveness of the network's recognition of the pump's condition. The whole study was summarized with conclusions. [ABSTRACT FROM AUTHOR]

Data in real world are usually characterized in multiple views, including different types of features or different modalities. Multi-view learning has been popular in the past decades and achieved significant improvements. In this paper, we investigate three challenging problems in the field of incomplete multi-view representation learning, namely, i) how to reduce the influences produced by missing views in multi-view dataset, ii) how to learn a consistent and informative representation among different views and iii) how to alleviate the impacts of the inherent noise in multi-view data caused by high-dimensional features or varied quality for different data points. To address these challenges, we integrate these three tasks into a problem and propose a novel framework termed Noise-aware Incomplete Multi-view Learning Networks (NIM-Nets). NIM-Nets fully utilize incomplete data from different views to produce a multi-view shared representation which is consistent, informative and robust to noise. We model the inherent noise in data by defining the distribution $\Gamma $ and assuming that each observation in the incomplete dataset is sampled from the distribution $\Gamma $. To the best of our knowledge, this is the first work to unify learning the consistent and informative representation, alleviating the impacts of noise in data and handling the view-missing patterns in multi-view learning into a framework. We also first give a definition of robustness and completeness for incomplete multi-view representation learning. Based on NIM-Nets, we present joint optimization models for classification and clustering, respectively. Extensive experiments on different datasets demonstrate the effectiveness of our method over the existing work based on classification and clustering tasks in terms of different metrics. [ABSTRACT FROM AUTHOR]

Social media play a significant role in communicating information across the globe, connecting with loved ones, getting the news, communicating ideas, etc. However, a group of people uses social media to spread fake information, which has a bad impact on society. Therefore, minimizing fake news and its detection are the two primary challenges that need to be addressed. This paper presents a multi-modal deep learning technique to address the above challenges. The proposed modal can use and process visual and textual features. Therefore, it has the ability to detect fake information from visual and textual data. We used EfficientNetB0 and a sentence transformer, respectively, for detecting counterfeit images and for textural learning. Feature embedding is performed at individual channels, whilst fusion is done at the last classification layer. The late fusion is applied intentionally to mitigate the noisy data that are generated by multi-modalities. Extensive experiments are conducted, and performance is evaluated against state-of-the-art methods. Three real-world benchmark datasets, such as MediaEval (Twitter), Weibo, and Fakeddit, are used for experimentation. Result reveals that the proposed modal outperformed the state-of-the-art methods and achieved an accuracy of 86.48%, 82.50%, and 88.80%, respectively, for MediaEval (Twitter), Weibo, and Fakeddit datasets. [ABSTRACT FROM AUTHOR]