Training Time Research Articles

Evaluation of the Performance of Neural and Non-Neural Methods to Classify the Severity of Work Accidents Occurring in the Footwear Industry Complex

In the footwear industry, occupational risks are significant, and work accidents are frequent. Professionals in the field prepare documents and reports about these accidents, but the need for more time and resources limits learning based on past incidents. Machine learning (ML) and deep learning (DL) methods have been applied to analyze data from these documents, identifying accident patterns and classifying the damage’s severity. However, evaluating the performance of these methods in different economic sectors is crucial. This study examined neural and non-neural methods for classifying the severity of workplace accidents in the footwear industry complex. The random forest (RF) and extreme gradient boosting (XGBoost) methods were the most effective non-neural methods. The neural methods 1D convolutional neural networks (1D-CNN) and bidirectional long short-term memory (Bi-LSTM) showed superior performance, with parameters above 98% and 99%, respectively, although with a longer training time. It is concluded that using these methods is viable for classifying accidents in the footwear industry. The methods can classify new accidents and simulate scenarios, demonstrating their adaptability and reliability in different economic sectors for accident prevention.

Diseño de un Sistema Integral para la Formación de Capital Humano en la pequeña y mediana empresa de Yucatán

The purpose of this study is to design an Integral System for the Formation of Human Capital (SIFCH), which allows the strengthening and specialization of this resource and, in this way, helps in the economic growth and consolidation of the market to which the products and services of small and medium-sized companies are directed. The method integrates the qualitative research approach with phenomenological processes, descriptive and analytical scope; for this purpose, interviews were conducted with 35 people, 17 businessmen of the State of Yucatan and 18 experts in the training area (consultants, private training companies, government agencies, business chambers, and business incubators). As a result, the Integral System of Human Capital Formation (SIFCH) was established, which consists of two parts. The internal part (the system's core) contemplates the organization's specific needs in terms of its size, business line, and resources; the entrepreneurial culture and innovation as an added value. The external part was designed according to a compendium of the elements that training experts consider at the time of training; these are: Diagnosis of the learning needs and the Evaluation of the Human Capital Training Process, that is, the results obtained in the productivity and personal growth of the worker in a period of time determined by the organization to measure the impact and effectiveness of this. The conclusions establish the processes to be followed in the implementation of SIFCH.

Image deduplication using efficient visual indexing and retrieval: optimizing storage, time and energy for deep neural network training

Image deduplication using efficient visual indexing and retrieval: optimizing storage, time and energy for deep neural network training

PSIII-12 Dynamics of neonatal dairy calf behaviors and their association with early gut microbiota

Abstract The behavior of dairy cattle can be both an indicator and a determinant of their productivity and health. Additionally, animal behavior is also associated with their gastrointestinal microbiota through the bi-directional pathways of the gut-brain axis, influencing nutrient utilization and gut health. Few studies have looked at the dynamics of neonatal calf behaviors and their associations with early microbiota development. Therefore, the objective of this study is to observe feeding behaviors and activity of neonatal dairy calves and investigate its associations with the diversity and composition of early-stage gut microbiota. We used Holstein-Angus crossed calves (n = 25) raised in individual outdoor hutches and bucket-fed milk replacer twice daily. The number of milk exposures until calves were successfully bucket trained was recorded, along with milk consumption times during the AM feedings on d 7, 11, and 14. We also monitored the activity of the calves using HOBO Pendant G Data Logger accelerometers secured on the right hind leg of each calf to monitor standing and lying behavior from d 3 to d 13. Fecal samples were collected from the recto-anal junction of the calves at d 7 and 14 after birth and gut microbiota were analyzed using full-length 16S rRNA gene amplicon sequencing on an Oxford Nanopore sequencer. The Generalized Linear Model (GLM) was applied to evaluate effects of gender and birth weight on behavior and gut microbiota using RStudio 4.2.1. The Linear discriminant analysis Effect Size (LEfSe) was used to determine the unique bacterial species for each behavioral group. The results revealed that there was no influence of sex or birth weight on bucket training or consumption time; however, male calves exhibited significantly increased total standing time compared with female calves. We found no significant difference in microbiota between fast-learning calves (one milk exposure; n = 3) and slow-learning calves (5 or more exposures; n = 3). We observed greater alpha diversity of slow drinking calves (top 25%; n = 6) compared with fast drinking calves (top 25%; n = 6) on d 7 (Chao1; P = 0.03). The LEfSe analysis identified Akkermansia muciniphilia to be more abundant in fast drinking calves, and Megasphaera elsdenii to be more abundant in slow drinking calves. Additionally, a trend of differing microbiota composition was observed between the most active and least active calves (n = 4 per group) on d 14 (Bray-Curtis; P = 0.1), and many Prevotella species were more abundant in the most active calves. In conclusion, this study sheds light on the gut-brain axis of neonatal dairy calves at the species-level and provides further insights into cognitive, consumptive, and active behaviors.

Surface tension prediction of pure organic components: An artificial neural network approach

Surface tension is an important thermodynamic property of fluids that impact several processes in the engineering. Accurate estimations of this parameter are required to correctly apply and develop operations such as flow processes and oil recovery. The present study proposes the applying of artificial neural network (ANN)-based approach to accurately predict the surface tension for 75 groups of pure organic compounds, encompassing 1,666 different compounds and 21,444 data points. The performance of the model was evaluated using the mean squared error (MSE) and the coefficient of determination (R2), with K-Fold cross-validation ensuring robustness. The optimal model (ArcD) achieved an average MSE of 8.27x10−5 during cross-validation and 1.55x10−5 during training, with an R2 value of 0.9955. Although training time for this model was 8.89 times longer than the second-best model (ArcA), which had an R2 of 0.9928. Comparative analyses with existing literature models were also realized and revealed an absolute relative deviation (ARD) value of 2.49% to ArcD, 18.39% to Brock and Bird, 23.97% to Curl and Pitzer, and 18.62% to Zuo and Stenby equations, demonstrating the applicability and precision of the developed model. The results highlighted the reliability, accuracy and robustness of the developed ANN-based model, providing a significant advancement in the prediction of surface tension for a wide range of pure organic compounds.

Performance comparison of machine learning algorithms for the estimation of blood pressure using photoplethysmography

This paper deals with an in-depth performance analysis on the estimation of Systolic Blood Pressure (SBP) and Diastolic Blood Pressure (DBP) by using features from the photoplethysmography (PPG) signal enhanced using the Maximal Overlap Discrete Wavelet Transform (MODWT), to train many machine learning (ML) regression models, including eXtreme Gradient Boost (XGBoost). The impact of different features selections methods, ML methods and training set sizes has been analyzed. One result is on the achievable improvements using features extracted from MODWT enhanced PPG signals. The most significant features have been selected using three different algorithms, namely RReliefF, Minimum Redundancy Maximum Relevance (MRMR) and Correlation-based Feature Selection (CFS). This comparison has been critical to underline that the exploitation of the new features allows to improve SBP and DBP estimation. Moreover, the authors have trained several ML algorithms to provide a comparison of their accuracy and training time, showing the Pareto frontier. RReliefF and MRMR selections algorithms, and several ML algorithms such as XGBoost, Gaussian Process Regression(GPR) and Ensemble stood out for their performance, with a different compromise between prediction error and training time. In addition, a further result has been obtained by varying the dimension of the dataset to understand the impact on Root Mean Square Error (RMSE) for models that have shown better performance, giving an empirical relationship on achievable RMSE as a function of training set size. From that relationship it has been extrapolated an upper boundary of the set size over which no further RMSE improvements are expected.

Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos: A Study of Neural Network Architectures

Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos: A Study of Neural Network Architectures

Virtual Reality as a Therapeutic Tool in Spinal Cord Injury Rehabilitation: A Comprehensive Evaluation and Systematic Review.

Introduction: The spinal rehabilitation process plays a crucial role in SCI patients' lives, and recent developments in VR have the potential to efficiently engage SCI patients in therapeutic activities and promote neuroplasticity. Objective: The primary objective of this study is to assess a complete review of the extended impacts of VR-assisted training on spine rehabilitation in SCI patients. Methods: This systematic review was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) through a single database search in PubMed/Medline between the dates 1 January 2010 and 1 February 2024. MESH terms and keywords were combined in the following search strategy: (Augmented Reality OR VR OR Virtual Reality) AND (Spine OR Spinal) AND Rehabilitation. Included articles were written in English, involved adults with SCI, included an intervention with VR, AR, or any mixed reality system, and assessed changes in outcomes after the intervention. Results: The search produced 257 articles, and 46 of them were allocated for data extraction to evaluate 652 patients. Both when VR training was analyzed and reviewed separately, and when compared to traditional training, the findings exhibited predominantly promising outcomes, reflecting a favorable trend in the study. VR technologies were used in different settings and customizations, and the medium total time of VR training among the studies was 60.46 h per patient. Conclusions: This auspicious outcome of the study further motivates the intervention of VR and AR in the rehabilitation of SCI patients along with ameliorating their overall holistic well-being.

Exploring the impact of impulsive urban sounds and noise on open sports training through real-time AI sound analysis

This study examines the impact of impulsive urban sounds and noise on athletes' performance during open sports training sessions. It specifically explores how various sports react to environmental noise disturbances. Impulsive urban sounds, defined as sudden, sharp noises that occur unpredictably in urban settings, were the focus of this study. Utilizing a sample of 160 sports students, divided into control and experimental groups, the study employed Independent Samples t-Tests and Analysis of Variance (ANOVA). These tests were used to assess athletic performance variations between concentration-intensive sports (e.g., archery, golf) and non-concentration-intensive sports (e.g., football, basketball). The findings reveal that while impulsive urban noises do not uniformly affect overall athletic performance, athletes in concentration-intensive sports experience a significant decline in athletic performance metrics, indicating a heightened vulnerability to such disturbances. In contrast, athletes in non-concentration-intensive sports showed a negligible impact from similar noise conditions. These observations suggest a complex interaction between the specific demands of a sport and its noise sensitivity, challenging the traditional view that urban noise universally disrupts athletic activities. The study highlights the importance of considering a sport's cognitive demands when evaluating the effects of environmental factors on athletic performance. It recommends strategic planning in the design and timing of urban sports training to minimize the negative impacts of impulsive noise, especially for disciplines that require intense concentration and mental focus. Future research should investigate individual adaptive capacities and develop customized noise mitigation strategies for different sports contexts. Keywords: urban noise, athletic performance, impulsive sounds, concentration-intensive sports, environmental stressors, noise mitigation, sports training environments.

U-DeepONet: U-Net enhanced deep operator network for geologic carbon sequestration

Learning operators with deep neural networks is an emerging paradigm for scientific computing. Deep Operator Network (DeepONet) is a modular operator learning framework that allows for flexibility in choosing the kind of neural network to be used in the trunk and/or branch of the DeepONet. This is beneficial as it has been shown many times that different types of problems require different kinds of network architectures for effective learning. In this work, we design an efficient neural operator based on the DeepONet architecture. We introduce U-Net enhanced DeepONet (U-DeepONet) for learning the solution operator of highly complex CO2-water two-phase flow in heterogeneous porous media. The U-DeepONet is more accurate in predicting gas saturation and pressure buildup than the state-of-the-art U-Net based Fourier Neural Operator (U-FNO) and the Fourier-enhanced Multiple-Input Operator (Fourier-MIONet) trained on the same dataset. Moreover, our U-DeepONet is significantly more efficient in training times than both the U-FNO (more than 18 times faster) and the Fourier-MIONet (more than 5 times faster), while consuming less computational resources. We also show that the U-DeepONet is more data efficient and better at generalization than both the U-FNO and the Fourier-MIONet.

Learning Autonomous Navigation in Unmapped and Unknown Environments.

Autonomous decision-making is a hallmark of intelligent mobile robots and an essential element of autonomous navigation. The challenge is to enable mobile robots to complete autonomous navigation tasks in environments with mapless or low-precision maps, relying solely on low-precision sensors. To address this, we have proposed an innovative autonomous navigation algorithm called PEEMEF-DARC. This algorithm consists of three parts: Double Actors Regularized Critics (DARC), a priority-based excellence experience data collection mechanism, and a multi-source experience fusion strategy mechanism. The algorithm is capable of performing autonomous navigation tasks in unmapped and unknown environments without maps or prior knowledge. This algorithm enables autonomous navigation in unmapped and unknown environments without the need for maps or prior knowledge. Our enhanced algorithm improves the agent's exploration capabilities and utilizes regularization to mitigate the overestimation of state-action values. Additionally, the priority-based excellence experience data collection module and the multi-source experience fusion strategy module significantly reduce training time. Experimental results demonstrate that the proposed method excels in navigating the unmapped and unknown, achieving effective navigation without relying on maps or precise localization.

Centralised rehearsal of decentralised cooperation: Multi-agent reinforcement learning for the scalable coordination of residential energy flexibility

This paper investigates the use of deep multi-agent reinforcement learning (MARL) for the coordination of residential energy flexibility. Particularly, we focus on achieving cooperation between homes in a way that is fully privacy-preserving, scalable, and that allows for the management of distribution network voltage constraints. Previous work demonstrated that MARL-based distributed control can be achieved with no sharing of personal data required during execution. However, previous cooperative MARL-based approaches impose an ever greater training computational burden as the size of the system increases, limiting scalability. Moreover, they do not manage their impact on distribution network constraints. We therefore adopt a deep multi-agent actor–critic method that uses a centralised but factored critic to rehearse coordination ahead of execution, such that homes can successfully cooperate at scale, with only first-order growth in computational time as the system size increases. Training times are thus 34 times shorter than with a previous state-of-the-art reinforcement learning approach without the factored critic for 30 homes. Moreover, experiments show that the cooperation of agents allows for a decrease of 47.2% in the likelihood of under-voltages. The results indicate that there is significant potential value for management of energy user bills, battery depreciation, and distribution network voltage management, with minimal information and communication infrastructure requirements, no interference with daily activities, and no sharing of personal data.

Bio-Plausible Multimodal Learning with Emerging Neuromorphic Devices.

Multimodal machine learning, as a prospective advancement in artificial intelligence, endeavors to emulate the brain's multimodal learning abilities with the objective to enhance interactions with humans. However, this approach requires simultaneous processing of diverse types of data, leading to increased model complexity, longer training times, and higher energy consumption. Multimodal neuromorphic devices have the capability to preprocess spatio-temporal information from various physical signals into unified electrical signals with high information density, thereby enabling more biologically plausible multimodal learning with low complexity and high energy-efficiency. Here, this work conducts a comparison between the expression of multimodal machine learning and multimodal neuromorphic computing, followed by an overview of the key characteristics associated with multimodal neuromorphic devices. The bio-plausible operational principles and the multimodal learning abilities of emerging devices are examined, which are classified into heterogeneous and homogeneous multimodal neuromorphic devices. Subsequently, this work provides a detailed description of the multimodal learning capabilities demonstrated by neuromorphic circuits and their respective applications. Finally, this work highlights the limitations and challenges of multimodal neuromorphic computing in order to hopefully provide insight into potential future research directions.

Improved Histogram of Oriented Gradient (HOG) Feature Extraction for Facial Expressions Classification

Facial expression classification system is one of the implementations of machine learning (ML) that takes facial expression datasets, undergoes training, and then utilizes the trained results to recognize facial expressions in new facial images. The recognized facial expressions include anger, contempt, disgust, fear, happy, sadness, and surprise expressions. The method employed for facial feature extraction utilizes histogram-oriented gradient (HOG). This study proposes an enhancement method for HOG feature extraction by reducing the feature dimension into multiple sub-features based on gradient orientation intervals, referred to as HOG channel (HOG-C). Classifier testing techniques are divided into two methods for comparison—support vector machines (SVM) with HOG features and SVM with HOG-C features. The testing results demonstrate that SVM with HOG achieves an accuracy of 99.9% with an average training time of 18.03 minutes, while SVM with HOG-C attains a 100% accuracy with an average training time of 18.09 minutes. The testing outcomes reveal that the implementation of SVM with HOG-C successfully enhances accuracy for facial expression classification.

DBSCAN SMOTE LSTM: Effective Strategies for Distributed Denial of Service Detection in Imbalanced Network Environments

In detecting Distributed Denial of Service (DDoS), deep learning faces challenges and difficulties such as high computational demands, long training times, and complex model interpretation. This research focuses on overcoming these challenges by proposing an effective strategy for detecting DDoS attacks in imbalanced network environments. This research employed DBSCAN and SMOTE to increase the class distribution of the dataset by allowing models using LSTM to learn time anomalies effectively when DDoS attacks occur. The experiments carried out revealed significant improvement in the performance of the LSTM model when integrated with DBSCAN and SMOTE. These include validation loss results of 0.048 for LSTM DBSCAN and SMOTE and 0.1943 for LSTM without DBSCAN and SMOTE, with accuracy of 99.50 and 97.50. Apart from that, there was an increase in the F1 score from 93.4% to 98.3%. This research proved that DBSCAN and SMOTE can be used as an effective strategy to improve model performance in detecting DDoS attacks on heterogeneous networks, as well as increasing model robustness and reliability.

Design, Building and Deployment of Smart Applications for Anomaly Detection and Failure Prediction in Industrial Use Cases

This paper presents a comparative analysis of deep learning techniques for anomaly detection and failure prediction. We explore various deep learning architectures on an IoT dataset, including recurrent neural networks (RNNs, LSTMs and GRUs), convolutional neural networks (CNNs) and transformers, to assess their effectiveness in anomaly detection and failure prediction. It was found that the hybrid transformer-GRU configuration delivers the highest accuracy, albeit at the cost of requiring the longest computational time for training. Furthermore, we employ explainability techniques to elucidate the decision-making processes of these black box models and evaluate their behaviour. By analysing the inner workings of the models, we aim at providing insights into the factors influencing failure predictions. Through comprehensive experimentation and analysis on sensor data collected from a water pump, this study contributes to the understanding of deep learning methodologies for anomaly detection and failure prediction and underscores the importance of model interpretability in critical applications such as prognostics and health management. Additionally, we specify the architecture for deploying these models in a real environment using the RAI4.0 metamodel, meant for designing, configuring and automatically deploying distributed stream-based industrial applications. Our findings will offer valuable guidance for practitioners seeking to deploy deep learning techniques effectively in predictive maintenance systems, facilitating informed decision-making and enhancing reliability and efficiency in industrial operations.

StyleRF-VolVis: Style Transfer of Neural Radiance Fields for Expressive Volume Visualization.

In volume visualization, visualization synthesis has attracted much attention due to its ability to generate novel visualizations without following the conventional rendering pipeline. However, existing solutions based on generative adversarial networks often require many training images and take significant training time. Still, issues such as low quality, consistency, and flexibility persist. This paper introduces StyleRF-VolVis, an innovative style transfer framework for expressive volume visualization (VolVis) via neural radiance field (NeRF). The expressiveness of StyleRF-VolVis is upheld by its ability to accurately separate the underlying scene geometry (i.e., content) and color appearance (i.e., style), conveniently modify color, opacity, and lighting of the original rendering while maintaining visual content consistency across the views, and effectively transfer arbitrary styles from reference images to the reconstructed 3D scene. To achieve these, we design a base NeRF model for scene geometry extraction, a palette color network to classify regions of the radiance field for photorealistic editing, and an unrestricted color network to lift the color palette constraint via knowledge distillation for non-photorealistic editing. We demonstrate the superior quality, consistency, and flexibility of StyleRF-VolVis by experimenting with various volume rendering scenes and reference images and comparing StyleRF-VolVis against other image-based (AdaIN), video-based (ReReVST), and NeRF-based (ARF and SNeRF) style rendering solutions.

A Fault Diagnosis Method for Electric Vehicle Lithium Power Batteries Based on Dual Feature Extraction from the Time and Frequency Domains

Abstract This study focuses on the safety and reliability issues of lithium-ion batteries, proposing a fault diagnosis strategy that leverages dual feature extraction in the time-frequency domains. Additionally, by modifying the traditional autoencoder, the study proposes a feature-guided autoencoder as an unsupervised model for extracting features in the time domain. Initially, wavelet packet decomposition and its energy denoising treatment are employed to refine fault information within battery voltage signals. Subsequently, the reconstruction error outputted by the Feature-Guided Autoencoder is utilized as the time-domain fault feature, while the cosine similarity of the energy of signals in various frequency bands obtained after wavelet packet decomposition serves as the frequency-domain fault feature. Ultimately, this paper selects the Isolation Forest algorithm for two-dimensional outlier detection of time-frequency features. Experimental results demonstrate that the Feature-Guided Autoencoder proposed in this study not only enhances the sensitivity of time-domain fault features compared to traditional autoencoders and their variants but also optimizes issues related to training time and computational load. The effectiveness of the proposed dual feature fault diagnosis method in both the time and frequency domains is validated through data from two actual vehicles, showing superior early fault detection capability relative to single-feature fault diagnosis methods.

Hand Surgery Training Structure in Germany - Evaluation of an Online Questionnaire

The duration of the training for hand surgery in Germany was reduced from three to two years in 2020, while other European Countries introduced hand surgery as a separate surgical specialty. Therefore, the structure of hand surgery training in Germany is under intense, ongoing discussion. This study aimed to evaluate the current situation of hand surgery training in Germany. Two separate online questionnaires were developed for self-assessment of residents and teachers in German hand surgery services. The questionnaire contained 29 questions for residents and 24 questions for teachers. Demographic data, educational tools, surgical training parameters and satisfaction were assessed from May to July 2023. The survey was based on the former training time of three years. The online questionnaire was sent to the members and associate members of the German Society for Hand Surgery via email. 148 participants answered the questionnaire, including 54 residents and 94 teachers. Overall, more female residents were assessed, whereas the majority of teachers was male. Hand surgery training is mostly conducted in high-level hospitals (tertiary care p=0.012, quaternary care p<0.001). Residents use digital teaching media significantly more often than teachers do (p<0.001). In contrast, teachers use journals (p=0.006) and clinical visits (p=0.024) significantly more often for their training than residents do. Trainers and trainees mostly agreed about time points of surgical milestones during surgical training that should be performed independently. Both trainers and trainees were largely satisfied with the practical and theoretical training in hand surgery. The importance of digital media in hand surgery training will increase in the near future. Therefore, digital media should be more frequently included in hand surgery training protocols. The study revealed that the majority of residents as well as teachers would appreciate the introduction of an independent hand surgery specialty, which would most likely improve the hand surgery competence in Germany.

The effect of 60days of 6° head-down-tilt bed rest on circulating adropin, irisin, retinol binding protein-4 (RBP4) and individual metabolic responses in young, healthy males.

Alterations in the circulating concentrations and target-tissue action of organokines underpin the development of insulin resistance in microgravity and gravity deprivation. The purpose of this study was to examine changes in circulating adropin, irisin, retinol binding protein-4 (RBP4), and the metabolic response of healthy young males following 60days of 6° head-down-tilt (HDT) bed rest, with and without reactive jump training (RJT), to explore links with whole-body and tissue-specific insulin sensitivity. To our knowledge, this is the first time that adropin, irisin, and RBP4 have been studied in HDT bed rest. A total of 23 male subjects (29 ± 6years, 181 ± 6cm, 77 ± 7kg) were exposed to 60days of 6° HDT bed rest and randomized to a control (CTRL, n = 11) or a RJT (JUMP, n = 12) group (48 sessions with ≤4min total training time per session). Circulating adropin, irisin, and RBP4 were quantified in fasting serum before and after HDT bed rest. A subanalysis was performed a posteriori to investigate individual metabolic responses post-HDT bed rest based on subjects that showed an increase or decrease in whole-body insulin sensitivity (Matsuda index). There were significant main effects of time, but not group, for decreases in adropin, irisin, Matsuda index, and liver insulin sensitivity following HDT bed rest (p < 0.05), whereas RBP4 did not change. The subanalysis identified that in a subgroup with decreased whole-body insulin sensitivity (n = 17), RBP4 increased significantly, whereas adropin, irisin, and liver insulin sensitivity were all decreased significantly following HDT bed rest. Conversely, in a subgroup with increased whole-body insulin sensitivity (n = 6), liver insulin sensitivity increased significantly after HDT bed rest, whereas adropin, irisin, and RBP4 did not change. Investigating individual metabolic responses has provided insights into changes in circulating adropin, irisin, RBP4, in relation to insulin sensitivity following HDT bed rest. We conclude that adropin, irisin, and RBP4 are candidate biomarkers for providing insights into whole-body and tissue-specific insulin sensitivity to track changes in physiological responsiveness to a gravity deprivation intervention in a lean male cohort.