Training Process Research Articles

Additional time error in association football is associated with interruption type and goal difference

ABSTRACT This study explores whether referees are biased when awarding additional time in Association Football. We collected data on Awarded Additional Time (AAT) granted by the referees and Calculated Additional Time (CAT), representing the interruption time meant to be considered for additional time according to the rulebook. Analysis of Additional Time Error (ATE), given as the difference between CAT and AAT, in the second half of German football Bundesliga season 2022/2023, revealed that referees award less AAT than warranted by interruptions (ATE = 2:10 ± 2:24 min). Multiple Linear Regression between either AAT or ATE and the duration of interruption types suggests that referees do not consider Goal Celebrations (for AAT: β = 0.09, p = 0.09, for ATE: β = 0.52, p < 0.01) as important as Substitutions (for AAT: β = 0.25, p < 0.01, for ATE: β = 0.03, p = 0.68), Injuries (for AAT: β = 0.35, p < 0.01, for ATE: β = 0.28, p < 0.01) and VAR (for AAT: β = 0.38, p < 0.01, for ATE: β = 0.10, p = 0.10). Mann-Whitney U-tests indicate a referee’s bias towards allocating insufficient AAT in matches with a goal difference greater than one compared (ATE = 3:16 ± 2:26 min) to the other matches (ATE = 1:18 ± 2:00 min) (p < 0.01, ES = 0.17). The Kruskal–Wallis H tests also show slight home bias from the referees, indicating that ATE is higher when the home team is leading more than one goal (ATE = 3:36 ± 2:36 min) (p < 0.01, ES = 0.15). In conclusion, although German Bundesliga referees adhere to the Laws of the Game, they show inconsistencies in handling different interruption types and matches with decisive outcomes. The findings suggest potential enhancement in referee training and decision-making processes and should be considered when advancing the game.

A neighborhood classifier based on adaptive radius selection and attribute reduction

Due to the capacity to process numerical data and generate highly interpretable results, neighborhood rough set theory has been widely used for data classification. However, there are two drawbacks that limit its performance. One is that the existing neighborhood classifiers suffer from insufficient information extraction due to the lack of a training process with feature selection. The other is that the neighborhood radius requires artificial parameters to adapt to different data distributions. To overcome the above shortcomings, a neighborhood classifier based on adaptive radius selection and attribute reduction (NRS-ASAR) is proposed. First, the K-nearest neighbor algorithm and the majority voting principle are introduced to eliminate noisy samples. Next, the training correlation radius is defined based on attribute reduction in which a relationship between the neighborhood radius and the features is established. Then, to overcome the subjectivity of the parameter in neighborhood classifier, an adaptive neighborhood radius called nearest neighborhood radius is defined. Finally, experimental results demonstrate that NRS-ASAR has better classification performance.

AED-PADA:Improving Generalizability of Adversarial Example Detection via Principal Adversarial Domain Adaptation

Adversarial example detection, which can be conveniently applied in many scenarios, is important in the area of adversarial defense. Unfortunately, existing detection methods suffer from poor generalization performance, because their training process usually relies on the examples generated from a single known adversarial attack and there exists a large discrepancy between the training and unseen testing adversarial examples. To address this issue, we propose a novel method, named Adversarial Example Detection via Principal Adversarial Domain Adaptation (AED-PADA). Specifically, our approach identifies the Principal Adversarial Domains (PADs), i.e., a combination of features of the adversarial examples generated by different attacks, which possesses a large portion of the entire adversarial feature space. Subsequently, we pioneer to exploit Multi-source Unsupervised Domain Adaptation in adversarial example detection, with PADs as the source domains. Experimental results demonstrate the superior generalization ability of our proposed AED-PADA. Note that this superiority is particularly achieved in challenging scenarios characterized by employing the minimal magnitude constraint for the perturbations.

Actas de órganos colegiados: nuevas disposiciones para el profesional en Administración de Oficinas

This article focuses on a review of the literature related to the new guidelines for the preparation of minutes. The main objective of this study is to analyze the legislative and regulatory changes implemented in Costa Rica, as well as the incorporation of technology in the management of collegiate bodies and the documentation associated with their work. The research is framed within the hermeneutic paradigm and was developed through a documentary review model that involved the compilation and thorough analysis of various bibliographic sources. Among the results found, it is highlighted that the legislative and regulatory changes have significantly transformed the dynamics of the drafting and format of this document, which requires a continuous training process for office administration professionals. One of the most relevant findings highlights the importance of administrative transparency for the proper management of organizations, in which the minutes play a crucial role as a document that evidences what was discussed and agreements made in the collegial bodies. It also emphasizes the existence of new provisions regulating the drafting and format of the minutes.

Challenge Acceptance as a Key Psychological Characteristic of Ukrainian Sergeants During the War: A Research Note

This exploratory pilot study aimed to identify psychological characteristics associated with combat readiness among Ukrainian sergeants engaged in the Russian–Ukrainian war. Eighty-three male sergeants (aged 26–60) from various combat units were assessed using psychological questionnaires and supervisor ratings of combat readiness. The findings revealed that challenge acceptance, a key component of hardiness, had the strongest positive correlation with combat readiness, while avoidance coping emerged as the most significant negative predictor. Additional negative coping strategies, such as indirect, antisocial, and aggressive actions, were also found to be negatively associated with readiness. These results provide initial insights into the psychological traits that may contribute to effective performance in combat and offer potential directions for refining selection and training processes for military personnel. Further research with larger samples is needed to confirm these findings and expand upon them.

Neural Radiance Fields with Hash-Low-Rank Decomposition

In recent advancements in novel view synthesis and neural rendering, neural radiance field (NeRF) has emerged as a powerful technique for synthesizing high-quality novel views of complex 3D scenes. However, the computational and storage demands of NeRF limit its applicability. In this paper, we present a novel approach to NeRF by combining low-rank decomposition and multi-hash encoding through a novel integration process to enhance efficiency and scalability. Our method reduces the model complexity and accelerates the training processes while maintaining high rendering quality. We demonstrate the effectiveness of our approach through extensive experiments on various datasets, showing significant improvements in performance and memory usage compared to traditional NeRF implementations. These results suggest that our approach can make NeRF more practical for real-world applications, such as virtual reality, gaming, and 3D reconstruction.

A New Approach to the Quality Determination of Used Palm Cooking Oil using Supervised Learning based on Electronic Sensors

As discarding used palm oil in nature is very dangerous, a processing mechanism is needed to utilize it according to needs. This utilization depends on the palm oil used, so the sorting process becomes important. This study proposes a new classification approach for the quality of used palm oil using Self-Organizing Map (SOM), Linear Vector Quantization (LVQ), and K-means, based on electronic sensors. This study included hardware design, software development, data collection, and training and testing processes. Based on the experimental results, the proposed system performed well using 13 parameters consisting of e-nose data, color, viscosity, and turbidity. The accuracy of SOM was 91.11%, LVQ achieved 95.56%, and K-Means obtained an accuracy of 98.89%. This system can be used as a decision support system in the automatic recognition of used palm oil to classify its quality.

Accurate prediction of colorectal cancer diagnosis using machine learning based on immunohistochemistry pathological images

Colorectal cancer (CRC) ranks as the third most prevalent tumor and the second leading cause of mortality. Early and accurate diagnosis holds significant importance in enhancing patient treatment and prognosis. Machine learning technology and bioinformatics have provided novel approaches for cancer diagnosis. This study aims to develop a CRC diagnostic model based on immunohistochemical staining image features using machine learning methods. Initially, CRC disease-specific genes were identified through bioinformatics analysis, SVM-RFE and Random Forest algorithm utilizing RNA-seq data from both GEO and TCGA databases. Subsequently, verification of these genes was performed using proteomics data from CPTAC and HPA database, resulting in identification of target proteins (AKR1B10, CA2, DHRS9, and ZG16) for further investigation. SVM and CNN were then employed to analyze and integrate the characteristics of immunohistochemical images to construct a reliable CRC diagnostic model. During the training and validation process of this model, cross-validation along with external validation methods were implemented to ensure accuracy and reliability. The results demonstrate that the established diagnostic model exhibits excellent performance in distinguishing between CRC and normal controls (accuracy rate: 0.999), thereby presenting potential prospects for clinical application. These findings are expected to provide innovative perspectives as well as methodologies for personalized diagnosis of CRC while offering more precise references for promising treatment.

Signs of consciousness in AI: Can GPT-3 tell how smart it really is?

The emergence of artificial intelligence (AI) is transforming how humans live and interact, raising both excitement and concerns—particularly about the potential for AI consciousness. For example, Google engineer Blake Lemoine suggested that the AI chatbot LaMDA might become sentient. At that time, GPT-3 was one of the most powerful publicly available language models, capable of simulating human reasoning to a certain extent. The notion of GPT-3 having some degree of consciousness could be linked to its ability to produce human-like responses, hinting at a basic level of understanding. To explore this further, we administered both objective and self-assessment tests of cognitive (CI) and emotional intelligence (EI) to GPT-3. Results showed that GPT-3 outperformed average humans on CI tests requiring the use and demonstration of acquired knowledge. However, its logical reasoning and EI capacities matched those of an average human. GPT-3’s self-assessments of CI and EI didn’t always align with its objective performance, with variations comparable to different human subsamples (e.g., high performers, males). A further discussion considered whether these results signal emerging subjectivity and self-awareness in AI. Future research should examine various language models to identify emergent properties of AI. The goal is not to discover machine consciousness itself, but to identify signs of its development, occurring independently of training and fine-tuning processes. If AI is to be further developed and widely deployed in human interactions, creating empathic AI that mimics human behavior is essential. The rapid advancement toward superintelligence requires continuous monitoring of AI’s human-like capabilities, particularly in general-purpose models, to ensure safety and alignment with human values.

“If I don’t do it, it feels wrong”: embodying environmental ethical reflexivity through cultivation and enactment of practices

Second-generation practice theoretical research on sustainable consumption has largely overlooked the role of environmental ethics. This is likely due to associations of “ethics” with culture and symbolic understandings of consumption. This article investigates how some environmentally aware and engaged consumers actively try to cultivate ethical and sustainable practices in order to develop ways of living more sustainably. By combining a practice theoretical framework with Michel Foucault’s work on ethical self-cultivation, I combine the ethical – understood as emerging in specific places and time as part of cultural formations – with processes of embodiment and ethical reflexivity to understand how environmental ethical reflexivity can play a role in the sustainable change of practices. The article argues that changing consumption practices based on changed environmental ethics requires an ongoing reflexive and bodily process of training the bodymind to embody skills and mold sensory, affective, and perceptual dispositions. This is a processual change requiring time and effort, in which research participants nevertheless experience meaning and hope as they are able to create sustainable change in their own lives and, over time and to some degree, live up to their ethical ideals. The analysis is based on ethnographic fieldwork conducted in two locations using audio-visual methods in participant observation and interviews with research participants who are actively engaged in changing their everyday practices.

Training improvement methods of ANN trajectory predictors in power systems

AbstractThis paper proposes training improvement methods of artificial neural networks (ANN) trajectory predictors. First, a dynamic power system time‐series trajectory is split into several different segments to simplify the original ANN training problem. Moreover, the time‐derivative of the trajectory is included to obtain an augmented loss function. Compared to previous studies which mainly focused on increasing the prediction accuracy, the aim of these novel techniques is to reduce the computational burden where the ANN output performance is still acceptable. The effectiveness of the developed methods is validated based on the WSCC three‐machine nine‐bus and IEEE 39‐bus system models. The mean absolute error (MAE) and trajectory prediction results are analysed, in which the numbers of neurons, hidden layers, and training epochs are constrained during the ANN training process. Rotor‐angle difference between generators and the system frequency are investigated as the dynamic trajectories of the power system models. The approaches are revealed to be effective when the ANN architecture and epochs are constrained. The MAE results can be reduced by up to 65% in the power system models depending on the ANN hyperparameters and training epochs. The ANN training results can better reflect the original trajectory as well.

A Study on Cyclical Learning Rates in Reinforcement Learning and Its Application to Temperature and Power Consumption Control of Refrigeration System

In recent years, with the advancement of computer hardware technology, an increasing number of complex control systems have begun employing reinforcement learning over traditional PID controls to address the challenge of managing multiple outputs simultaneously. In this study, we have for the first time adopted the cyclical learning rate method, which is widely used in deep learning, and applied it to deep reinforcement learning. Utilizing MATLAB Simulink, a detailed simulation model was developed with the RSD-4TFK5J model refrigeration storage as the reference object. We precisely evaluated the effects of various cyclical learning rate strategies on the training process of the model. The simulation results demonstrate the effectiveness of the cyclical learning rate method during the training phase, showcasing its potential to enhance learning efficiency and system performance in complex control environments.

HC 2L: Hybrid and Cooperative Contrastive Learning for Cross-Lingual Spoken Language Understanding.

State-of-the-art model for zero-shot cross-lingual spoken language understanding performs cross-lingual unsupervised contrastive learning to achieve the label-agnostic semantic alignment between each utterance and its code-switched data. However, it ignores the precious intent/slot labels, whose label information is promising to help capture the label-aware semantics structure and then leverage supervised contrastive learning to improve both source and target languages' semantics. In this paper, we propose Hybrid and Cooperative Contrastive Learning to address this problem. Apart from cross-lingual unsupervised contrastive learning, we design a holistic approach that exploits source language supervised contrastive learning, cross-lingual supervised contrastive learning and multilingual supervised contrastive learning to perform label-aware semantics alignments in a comprehensive manner. Each kind of supervised contrastive learning mechanism includes both single-task and joint-task scenarios. In our model, one contrastive learning mechanism's input is enhanced by others. Thus the total four contrastive learning mechanisms are cooperative to learn more consistent and discriminative representations in the virtuous cycle during the training process. Experiments show that our model obtains consistent improvements over 9 languages, achieving new state-of-the-art performance.

Robust Semi-Supervised Learning by Wisely Leveraging Open-Set Data.

Open-set Semi-supervised Learning (OSSL) holds a realistic setting that unlabeled data may come from classes unseen in the labeled set, i.e., out-of-distribution (OOD) data, which could cause performance degradation in conventional SSL models. To handle this issue, except for the traditional in-distribution (ID) classifier, some existing OSSL approaches employ an extra OOD detection module to avoid the potential negative impact of the OOD data. Nevertheless, these approaches typically employ the entire set of open-set data during their training process, which may contain data unfriendly to the OSSL task that can negatively influence the model performance. This inspires us to develop a robust open-set data selection strategy for OSSL. Through a theoretical understanding from the perspective of learning theory, we propose Wise Open-set Semi-supervised Learning (WiseOpen), a generic OSSL framework that selectively leverages the open-set data for training the model. By applying a gradient-variance-based selection mechanism, WiseOpen exploits a friendly subset instead of the whole open-set dataset to enhance the model's capability of ID classification. Moreover, to reduce the computational expense, we also propose two practical variants of WiseOpen by adopting low-frequency update and loss-based selection respectively. Extensive experiments demonstrate the effectiveness of WiseOpen in comparison with the state-of-the-art.

Icon similarity model based on cognition and deep learning

Human-computer cooperation guided by natural interaction, intelligent interaction, and human–computer integration is gradually becoming a new trend in human–computer interfaces. An icon is an indispensable pictographic symbol in an interface that can convey pivotal semantics between humans and computers. Research on similar icons’ cognition in humans and the discrimination of computers can reduce misunderstandings and facilitate transparent cooperation. Therefore, this research focuses on images of icons, extracted contours, and four features, including the curvature, proportion, orientation, and line of the contour, step by step. By manipulating the feature value change to obtain 360 similar icons, a cognitive experiment was conducted with 25 participants to explore the boundary values of the feature dimensions that cause different levels of similarity. Its boundary values were applied to deep learning to train a discrimination algorithm model that included 1500 similar icons. This dataset was used to train a Siamese neural network using a 16-layer network branch of a visual geometry group. The training process used stochastic gradient descent. This method of combining human cognition and deep learning technology is meaningful for establishing a consensus on icon semantics, including content and emotions, by outputting similarity levels and values. Taking icon similarity discrimination as an example, this study explored the analysis and simulation methods of computer vision for human visual cognition. The accuracy evaluated is 90.82%. The precision was evaluated as 90% for high, 80.65% for medium, and 97.30% for low. Recall was evaluated as 100% for high, 89.29% for medium, and 83.72% for low. It has been verified that it can compensate for fuzzy cognition in humans and enable computers to cooperate efficiently.

PKDGAN: Private Knowledge Distillation With Generative Adversarial Networks

The deployment of deep learning applications has to address the increasing privacy concerns when using private and sensitive data for training. A conventional deep learning model is prone to privacy attacks that can recover the sensitive information from either model parameters or accesses to the inference model. Recently, differential privacy (DP) has been proposed to offer provable privacy guarantees by randomizing the training process of neural networks. However, many approaches tend to provide the worst case privacy protection for model publishing, inevitably impairing the accuracy of the trained models. Thus, we present a novel private knowledge transfer strategy, where the private teacher trained on sensitive data is not publicly accessible but the student models can be released with privacy guarantees. In this paper, a three-player (teacher-student-discriminator) learning framework, Private Knowledge Distillation with Generative Adversarial Networks (PKDGAN), is proposed, where the student acquires the distilled knowledge from the teacher and is trained with the discriminator to generate similar outputs as the teacher. Moreover, a cooperative learning strategy is also suggested to support the collective training of multiple students against the discriminator when each student is with insufficient unlabelled training data. To enforce rigorous privacy guarantees, PKDGAN applies a Rényi differential privacy mechanism throughout the training process, and use it with a moment accountant technique to track the privacy cost. PKDGAN allows students to be trained with unlabelled public data and very few epochs, which avoids the exposure of training data while ensuring model performance. In the experiments, PKDGAN is found to have consistently good performance on various datasets (MNIST, SVHN, CIFAR-10, and Market-1501). When compared to prior works [1], [2], PKDGAN exhibits 5-82% accuracy loss improvement without compromising any privacy guarantee.

Safe deep reinforcement learning-assisted two-stage energy management for active power distribution networks with hydrogen fueling stations

In a power-hydrogen coupled integrated energy system (PHCIES), hydrogen fueling stations (HFSs) with solar photovoltaic (PV) systems are crucial devices to support hydrogen demand and maintain a stable and near-zero pollutant emission-based PHCIES operation by producing a pollutant-free hydrogen. However, optimal operation management of HFSs is challenging because the electrolyzers, compressor, hydrogen storage system (HSS), and fuel cell in HFSs are interconnected and change dynamically under various operational environments. To resolve this issue, this paper presents a safe deep reinforcement learning (DRL)-assisted two-stage framework that ensures a reliable and economical PHCIES operation. In the first stage, day-ahead operational schedules of stand-alone PV systems, PV-enabled HFSs, on-load tap changers, and capacitor banks with hourly resolution are coordinated to minimize the system operation cost, electricity arbitrage cost, PV curtailment cost, and real power loss by solving a Volt-VAR control (VVC) optimization problem. In the second stage, a safe DRL algorithm with a 15-min resolution is employed to minimize the real power loss and maximize the HFS profit by rescheduling the reactive power of stand-alone PV systems and real/reactive power and hydrogen of HFSs. The proposed self-tuning adaptive safety module integrated in the DRL method ensures no violations of HSS’ state-of-hydrogen (SOH) and voltage magnitude in the PHCIES during the training process. Numerical examples conducted on a PHCIES (IEEE 33-bus system coupled with two PV-enabled HFSs) demonstrate the effectiveness of the proposed framework in terms of training convergence, SOH/voltage violation, real power loss, and profit of HFSs.

Gradient Inversion Attacks: Impact Factors Analyses and Privacy Enhancement.

Gradient inversion attacks (GIAs) have posed significant challenges to the emerging paradigm of distributed learning, which aims to reconstruct the private training data of clients (participating parties in distributed training) through the shared parameters. For counteracting GIAs, a large number of privacy-preserving methods for distributed learning scenario have emerged. However, these methods have significant limitations, either compromising the usability of global model or consuming substantial additional computational resources. Furthermore, despite the extensive efforts dedicated to defense methods, the underlying causes of data leakage in distributed learning still have not been thoroughly investigated. Therefore, this paper tries to reveal the potential reasons behind the successful implementation of existing GIAs, explore variations in the robustness of models against GIAs during the training process, and investigate the impact of different model structures on attack performance. After these explorations and analyses, this paper propose a plug-and-play GIAs defense method, which augments the training data by a designed vicinal distribution. Sufficient empirical experiments demonstrate that this easy-to-implement method can ensure the basic level of privacy without compromising the usability of global model.

Critical analysis of information provided by ChatGPT on lactate, exercise, fatigue, and muscle pain: current insights and future prospects for enhancement.

This study aimed to critically evaluate the information provided by ChatGPT on the role of lactate in fatigue and muscle pain during physical exercise. We inserted the prompt "What is the cause of fatigue and pain during exercise?" using ChatGPT versions 3.5 and 4o. In both versions, ChatGPT associated muscle fatigue with glycogen depletion and "lactic acid" accumulation, whereas pain was linked to processes such as inflammation and microtrauma. We deepened the investigation with ChatGPT 3.5, implementing user feedback to question the accuracy of the information about lactate. The response was then reformulated, involving a scientific debate about the true role of lactate in physical exercise and debunking the idea that it is the primary cause of muscle fatigue and pain. We also utilized the creation of a "well-crafted prompt," which included persona identification and thematic characterization, resulting in much more accurate information in both the ChatGPT 3.5 and 4o models, presenting a range of information from the physiological process of lactate to its true role in physical exercise. The results indicated that the accuracy of the responses provided by ChatGPT can vary depending on the data available in its database and, more importantly, on how the question is formulated. Therefore, it is indispensable that educators guide their students in the processes of managing the AI tool to mitigate risks of misinformation.NEW & NOTEWORTHY Generative artificial intelligence (AI), exemplified by ChatGPT, provides immediate and easily accessible answers about lactate and exercise. However, the reliability of this information may fluctuate, contingent upon the scope and intricacy of the knowledge derived from the training process before most recent update. Furthermore, a deep understanding of the basic principles of human physiology becomes crucial for the effective correction and safe use of this technology.

The application of “transfer learning” in optical microscopy: The petrographic classification of opaque minerals

Abstract The analysis of optical microscopic image data is crucial for the identification and characterization of mineral phases and, thus, directly relevant to the subsequent methodology selections of further detailed petrological exploration. Here, we present a novel application of Swin Transformer, a deep learning algorithm to classify mineral phases such as arsenopyrite, chalcopyrite, gold, pyrite, and stibnite in images captured by optical microscopy. To speed up the training process and improve the generalization capabilities of the investigated model, we adopt the “transfer learning” paradigm by pre-training the algorithm using a large, general-purpose image data set named ImageNet-1k. Furthermore, we compare the performances of the Swin Transformer with those of two well-established Convolutional Neural Networks (CNNs) named MobileNetv2 and ResNet50, respectively. Our results highlight a maximum accuracy of 0.92 for the Swin Transformer, outperforming the CNNs. To provide an interpretation of the trained models, we apply the so-called Class Activation Map (CAM), which indicates a strong global feature extraction ability of the Swin Transformer metal mineral classifier that focuses on distinctive (e.g., colors) and microstructural (e.g., edge shapes) features. The results demonstrate that the deep learning approach can accurately extract all available attributes, which reveals the potential to assist in data exploration and provides an opportunity to carry out spatial quantization at a large scale (centimeters-millimeters). Simultaneously, boosting the learning processes with pre-trained weights can accurately capture relevant attributes in mineral classification, revealing the potential for application in mineralogy and petrology, as well as enabling its use in resource explorations.