Efficient Training Research Articles

The Investigation of Hyperparameters Influence on Fruit and Vegetable Image Recognition Performance Using SVM

This study aims to improve the efficiency of automatic classification and quality control of fruits and vegetables through image recognition technology, to achieve efficient and accurate intelligent sorting in agricultural production, reduce labor costs and improve market competitiveness. A high-quality image dataset of 36 fruit and vegetable categories from Kaggle is used in this study. The images in the dataset have been preprocessed to ensure that the data is suitable for the classification task and sets the stage for efficient training and evaluation of the model. Logistic regression was first used as the baseline model in order to compare the performance with the Support Vector Machine (SVM) model. Subsequently, hyperparameter tuning is performed to optimize the model to achieve the best cross-validation accuracy. Next, the SVM model is trained with the selected hyperparameters, and the training time is recorded. The performance of the model was evaluated in detail by confusion matrix and classification reports, and the test set was used for final validation to ensure that the model would also perform well on unseen data. The SVM model achieves an accuracy of 96% on both validation and test sets, which is a very good performance. The hyperparameters optimized by GridSearchCV (C = 10, gamma = 0.1, kernel = rbf) effectively improve the performance of the model, verifying that reasonable hyperparameter selection is crucial to the SVM model. These results show that the model has good generalization ability and potential to be applied to real-time classification tasks.

Isokinetic Force Modeling in Sports Based on Embedded System of Motion Recognition Model

With the gradual deepening of sports research, sports training methods integrating more scientific and technological means can greatly improve athletes’ professional ability. Combined with the isokinetic strength training method, it can effectively alleviate muscle recovery and body flexibility, which requires an analysis of the level of sports body promotion, and a more efficient training program for various sports parameters. In this study, taking strength enhancement as an example, the convolutional neural network is innovatively integrated into isokinetic strength training. Using the motion data obtained by wearable sensors in the training, we trained the neural network model by real-time analysis of data samples through an embedded system. Further, we optimized the network parameters to realize the extraction and recognition of EMG depth features. Compared with the traditional isokinetic strength classification model, this model lays a foundation for further constructing personalized strength training guidance methods. After the simulation test, the difference of the model’s accuracy requirement before and after the pre-training is 4.6%, and the difference of the test set before and after the pre-intensive training is [Formula: see text]1%. It is 16.2% higher than the TCAM algorithm, and the gap with STA-LSTM remains between [Formula: see text]. It is verified that the attention concentration control scheme can not only reduce the error rate in visual recognition tasks, but also improve the complexity of the neural network model. The optimized lightweight data structure is especially suitable for embedded systems with limited computing power. All kinds of test projects verify its rationality and make full use of the control ability of the embedded system and the characteristics of parallel computing and programmability, which has great practical application value for the development of the Internet of Things system network. The research results have made a thorough study of the physiological guidance of existing strength training, the identification and evaluation methods of exercise-induced muscle fatigue, and the current situation of the existing isokinetic equipment.

Equilibrio y carga plantar en diferentes técnicas de sentadilla

This research aims to analyze dynamic balance and plantar load in different squatting techniques, providing insights into the influence of these exercises in practitioners who incorporate weightlifting into their training routines. A sample of 21 subjects engaged in activities involving three squatting techniques: Back Squat (BS), Front Squat (FS), and Overhead Squat (OHS) was analyzed. Participants assumed a baseline posture and performed exercises with approximately 50% RM for each squatting technique. A plantar pressure measurement system was used to assess center of pressure (COP) displacement, peak pressure (PP), and plantar load in different regions of the foot. One-way Welch's ANOVA considering COP displacement, PP, and percentage of force for each foot region was measured. The results showed no significant effect of the squatting technique on COP trajectory during the exercises. However, a significant interaction was observed between the squatting technique and PP compared to the static balance situation. All three squats resulted in higher PP than static balance, with OHS exhibiting the highest values, followed by FS and BS. No significant differences were found between the squat techniques. These findings suggest that different squat techniques can influence plantar load, but do not affect the trajectory of the COP during the execution of the exercise. The heel region generated the greatest plantar load among the three squat techniques and should receive greater attention during the squat and its variations. This information can assist in prescribing exercises and developing more efficient training programs for weightlifters.

Boosting the reconstruction performance of 3D Multi-porous media using double generative adversarial networks

With the continuous improvement of mathematical modeling technology, reconstructing the three-dimensional structure of media from two-dimensional reference images has become an important research method for the three-dimensional modeling of multi-porous media. Deep-learning-based methods are currently popular and form the focus of this research field. However, the performance of deep learning in reconstructing the three-dimensional structure of media from two-dimensional reference images still requires improvement. To enhance the diversity and generalization of network-generated three-dimensional models, this study proposed a preprocessing method that correlated two-dimensional reference images with Gaussian noise, a three-orthogonal random section constraint method, and a dual generative adversarial network (DGAN)-based model. Multiple sets of core samples, a set of building materials, and a set of battery-material samples were used to verify the performance of the proposed network. Both intuitive morphological and statistical feature comparisons showed that the DGAN model solved the problem of insufficient diversity and generalization when reconstructing three-dimensional porous media from a single image using deep-learning-based methods. The morphological and statistical features of the reconstructed three-dimensional structure also exhibited good consistency with the reference two-dimensional image, and the training efficiency of the network was greatly improved.

Application of Intelligent Collaboration Technology in Skill Training and Job Docking of Data Tagger

With the rapid development of artificial intelligence and machine learning technology, the role of data tagger has become more and more important, responsible for providing high-quality annotated data for training algorithms, but traditional skills training and job docking methods cannot meet the needs of the rapid development of the industry. This paper studies the application of intelligent collaboration technology in improving the skill training efficiency of data taggers and promoting its effective docking with the post. By using the literature review method, in-depth interview method, and case analysis method, this paper analyzes the current challenges faced by data taggers, including skill gaps, insufficient training resources, and rapid changes in market demand. The results show that the application of intelligent collaboration technology can not only improve the skill training efficiency of data taggers but also promote effective docking between them and their positions, bringing a positive impact to the data tagging industry.

A framework for measuring the training efficiency of a neural architecture

Measuring Efficiency in neural network system development is an open research problem. This paper presents an experimental framework to measure the training efficiency of a neural architecture. To demonstrate our approach, we analyze the training efficiency of Convolutional Neural Networks and Bayesian equivalents on the MNIST and CIFAR-10 tasks. Our results show that training efficiency decays as training progresses and varies across different stopping criteria for a given neural model and learning task. We also find a non-linear relationship between training stopping criteria, training Efficiency, model size, and training Efficiency. Furthermore, we illustrate the potential confounding effects of overtraining on measuring the training efficiency of a neural architecture. Regarding relative training efficiency across different architectures, our results indicate that CNNs are more efficient than BCNNs on both datasets. More generally, as a learning task becomes more complex, the relative difference in training efficiency between different architectures becomes more pronounced.

Tell me and I forget, teach me and I may remember, involve me and I learn: impact of simulation training on performance in electrophysiology

Abstract Background Interpretation of intracardiac electrograms plays a key role in electrophysiology (EP) training. However, there had been a substantial gap between theoretical knowledge and practical skills acquired in the clinical setting. With the development of an online life-like EP-simulator, interactive training has never been so easily accessible. Purpose The aim of this study is to assess the efficiency of simulation-based training in electrophysiology, with the help of EP-simulator, by comparing it to traditional-learning methods. Methods 340 participants attended a 3-hour, online video-lecture on basic electrophysiology. Then, they were randomised 1:1 into a traditional-learning and simulation-learning groups. Both groups received presentation slides from the lecture and two textbook recommendations. In addition, simulation-learning group obtained access to online EP-simulator. This was followed by one week of self-learning. At the end, their knowledge was assessed using an online multiple-choice test. Results Eventually, 91 participants from the traditional-learning group and 67 participants from the simulation-learning group took the final test. The simulation-learning group, which obtained additionally an access to EP-simulator, scored significantly better than the traditional-learning group (mean score=15.10 vs 13.16, p=0.022). Conclusion EP-simulator can possibly contribute to better learning results. Including simulation in training programs may have an impact on educational outcomes.Course of the studyEP-simulator interface

Research on frequency shift signal detection algorithm for jointless track circuits based on improved relaxation algorithm

The accuracy of ZPW-2000 frequency shift signal demodulation is related to the safety and efficiency of high-speed trains. This paper proposed a kind of ZPW-2000 frequency-shift signal detection algorithm based on improved Relaxation algorithm (ZFSD-IR) to address the issue of low accuracy in detecting ZPW-2000 frequency shift signal under strong noise interference, as the traditional fast Fourier transform (FFT) spectrum analysis methods are susceptible to fence effects and spectrum leakage effects. Firstly, the principle of ZPW-2000 jointless track circuit was analyzed, and then the ideal and simplified models of ZPW-2000 frequency shift signal were established, respectively. Finally, the simulation experiments were conducted on ZPW-2000 frequency shift signal detection under different sampling durations and signal-to-noise ratios (SNR). The effectiveness of the proposed algorithm was verified through comparative analysis with the traditional algorithms. The research results indicated that the ZFSD-IR algorithm has better accuracy in carrier-frequency and low-frequency estimation than traditional algorithms, and can achieve accurate detection of ZPW-2000 frequency shift signal under the lower SNR conditions. It has high detection accuracy and good anti-interference ability, ensuring the safe operation of high-speed trains.

Challenges of Lack of Human Resources in Training Management at SPNF Sanggar Medan City Learning Activities

The background of the research is based on the formulation of the problem, namely the challenges faced by the Learning Activity Studio (SKB) in Medan City from the perspective of training management due to limited resources. This study uses a qualitative approach with a descriptive analysis method. The subjects in this study are leaders and employees at SPNF SDKB in Medan City who are selected through purposive sampling techniques. Data collection was carried out through observation and interviews directly to respondents. Data analysis is carried out qualitatively consisting of data reduction, data presentation, and conclusion drawing. The results of the analysis show that the development of human resources in a SKB is very necessary to improve the quality and success of the training program, namely further efforts are needed in the development of more varied training programs and more activities, as well as improving human resource competencies so that the company can survive and develop in accordance with environmental demands. To increase training management resources, SKB can consider strategies such as internal and external training, as well as the use of technology to improve training access and efficiency.

Impact of Smartphone-Induced Mental Fatigue on Resistance Training Performance and Efficiency - The Role of Smartphone Use During Workout

Introduction Resistance training volume is a key determinant of muscle hypertrophy and strength development. However, mental fatigue can significantly impact this volume, ultimately affecting performance. One prominent source of mental fatigue in modern gym environments is smartphone use, which has become pervasive and may reduce workout efficiency and overall training volume. Purpose This review examines the dual effects of smartphone use on resistance training outcomes and offers strategies to balance these impacts effectively. Conclusions Smartphone-induced mental fatigue increases perceived exertion, leading to reduced performance during resistance training by lowering volume-load and endurance, particularly in low- and moderate-intensity exercises. Despite this, maximal strength appears to remain unaffected. Additionally, smartphone use diverts cognitive resources, resulting in mental fatigue and impairing both physical and technical performance due to distractions during workouts. It can also unintentionally prolong rest periods, which disrupts the balance between muscle recovery and mental readiness, further decreasing training efficiency. However, smartphones also offer valuable benefits, such as access to fitness apps, progress tracking, and motivational music, which can help maintain or even enhance performance by improving focus and motivation. To minimize distractions and optimize training outcomes, strategies such as using airplane mode or scheduling specific times for phone use are recommended. While excessive smartphone use can impair training efficiency, careful management and strategic use can enable athletes to capitalize on its benefits and improve resistance training results. Future research should focus on developing best practices for incorporating smartphone use in fitness settings to balance its advantages while minimizing its downsides.

A multi-step regularity assessment and joint prediction system for ordering time series based on entropy and deep learning

Customer maintenance is of vital importance to the enterprise management. Valuable assessment and efficient prediction for customer ordering behavior can offer better decision-making and reduce business costs significantly. According to existing studies about customer behavior regularity segment and demand prediction most focus on e-commerce and other fields with large amount of data, making them not suitable for small enterprises and data features like sparsity and outliers are not mined when doing regularity quantification. Additionally, more and more complex network structures for demand prediction are proposed, which builds on the assumption that all the samples have predictive value, ignoring the fine-grained analysis of different time series regularity with high cost. To deal with the above issues, a multi-step regularity assessment and joint prediction system for ordering time series is proposed. For extracting features, comprehensive assessment of customer regularity based on entropy weight method with the result of predictability quantification using K-Means clustering algorithm, real entropy, LZW algorithm and anomaly detection adopting Isolation Forest algorithm not only gives an objective result to ‘how high the regularity of customers is’, filling the gap in the field of regularity quantification, but also provides a theoretical basis for demand prediction models selection. Prediction models: Random Forest regression, XGBoost, CNN and LSTM network are experimented with sMAPE and MSLE for performance evaluation to verify the effectiveness of the proposed regularity quantitation method. Moreover, a merged CNN-BiLSTM neural network model is established for predicting those customers with low regularity and difficult to predict by traditional machine leaning algorithms, which performs better on the data set compared to others. Random Forest is still used for prediction of customers with high regularity due to its high training efficiency. Finally, the results of prediction, regularity quantification, and classification are output from the intelligent system, which is capable of providing scientific basis for corporate strategy decision and has highly extendibility in other enterprises and fields for follow-up research.

SIAT: Document-level Event Extraction via Spatiality-Augmented Interaction Model with Adaptive Thresholding

Document-level event extraction endeavors to automatically extract structural events from a given document. Many existing approaches focus on modeling entity interactions and decoding these interactions into events, assigning each entity as an event argument. However, these approaches encounter two primary limitations: they exclusively capture semantic dependencies to model entity interactions, overlooking the indication of the spatial distribution features of entities; they decode interactions imprecisely with a hard binary-classification boundary, potentially failing to calibrate micro differences in interactions. To overcome these limitations, we introduce a novel approach termed the S patiality-augmented I nteraction Model with A daptive T hresholding (SIAT). Our method addresses the first limitation by calculating the relative position encoding of entities to represent spatial interaction features. These features are then integrated with multi-granularity semantic interactions, enhancing the modeling of entity interactions for each entity pair. Furthermore, we introduce an adaptive event decoding mechanism, which establishes a more flexible decision boundary for different entity interactions. Additionally, an adaptive loss function for threshold learning is designed to further refine the model. Experimental results demonstrate that our proposed method achieves competitive performance compared to state-of-the-art methods on two public event extraction datasets while maintaining considerable training efficiency.

Repeated-sprint training in hypoxia: A review with 10 years of perspective.

Over the past decade, numerous studies have investigated an innovative "live low-train high" approach based on the repetition of short (<30 s) "all-out" sprints with incomplete recoveries in hypoxia; the so-called Repeated-Sprint training in Hypoxia (RSH). The aims of the present review are therefore threefold. First, this study summarizes the available evidence on putative additional performance enhancement after RSH comparing to the same training in normoxia (RSN). Second, a critical analysis of underpinning mechanisms discusses how advantages can be obtained through RSH for sea-level performance enhancement. An enhanced microcirculatory vasodilation leading to improved muscle perfusion and/or oxygenation and an increase in muscular phosphocreatine content may help explain the superiority of RSH vs. RSN. Third, the present review aims to provide guidelines for coaches, athletes and scientists to apply RSH interventions with regard to the interval duration, exercise-to-rest ratio and training volume. In conclusion, this review supports repeated-sprint training in hypoxia as an efficient (but not magic) training intervention with 77% of the controlled studies reporting an additional benefit with added hypoxia, mainly for team-, combat- and racket-sports athletes but also for all other sports (e.g. endurance) that require repeated accelerations with lesser fatigue.

Post-earthquake inspection of high-speed railway viaducts with multi-scale task interaction deep learning strategy

Automated computer vision-based inspections of railway infrastructures, such as component type, damaged status, and location, have been investigated actively by resorting to task-specific deep learning models. However, task-specific models that fulfill these separate inspection tasks encountered bottlenecks in improving inference accuracy, and bring huge computational costs. Multi-task deep learning, which can fulfill these inspection tasks concurrently, has yet to be fully investigated in the context of structural inspection. In this study, a multi-scale task interaction deep learning strategy is presented towards component recognition, damage segmentation, and depth estimation for a comprehensive post-earthquake inspection of high-speed railway viaducts. Three modules for multi-scale task interaction were proposed to modify the multi-task deep neural network, taking full advantage of task commonalities at multiple scales. The proposed method was validated with a large-scale image dataset of high-speed railway viaducts. Component recognition and depth estimation were incorporated to implement multi-task learning since they have higher pattern affinities at multiple scales. Results reported that mean Intersection over Unions of testing samples of component and damage tasks were 91.2% and 72.1%, RMSE of depth estimation was 1.54 m. Compared with single-task cases, training time, inference duration, and FLOPs of the multi-task model were reduced by 23%, 30%, 27% respectively. Results showed improvement in both inference accuracy and training efficiency, substantiating the superiority of the proposed strategy.

The Application of Augmented Reality (AR) and Virtual Reality (VR) in Apparel Manufacturing and Operator Training

The integration of Augmented Reality (AR) and Virtual Reality (VR) into various industries has revolutionized training, production processes, and operational efficiency. The apparel industry faces challenges in operator training, resulting in decreased productivity and increased errors when frequent style change occurs. The creation of an AR/VR system for training purposes could serve as a valuable tool. In the sewing and apparel industry, these technologies have been gaining attention for their potential to streamline processes, enhance skill development, and optimize operator training. This study contributes to the apparel industry by introducing innovative education and training methods, enhancing the skills and knowledge of professionals, and improving manufacturing efficiency and quality and the other broader benefits these technologies offer.

Forecasting trends in food security with real time data

Early warning systems are an essential tool for effective humanitarian action. Advance warnings on impending disasters facilitate timely and targeted response which help save lives and livelihoods. In this work we present a quantitative methodology to forecast levels of food consumption for 60 consecutive days, at the sub-national level, in four countries: Mali, Nigeria, Syria, and Yemen. The methodology is built on publicly available data from the World Food Programme’s global hunger monitoring system which collects, processes, and displays daily updates on key food security metrics, conflict, weather events, and other drivers of food insecurity. In this study we assessed the performance of various models including Autoregressive Integrated Moving Average (ARIMA), Extreme Gradient Boosting (XGBoost), Long Short Term Memory (LSTM) Network, Convolutional Neural Network (CNN), and Reservoir Computing (RC), by comparing their Root Mean Squared Error (RMSE) metrics. Our findings highlight Reservoir Computing as a particularly well-suited model in the field of food security given both its notable resistance to over-fitting on limited data samples and its efficient training capabilities. The methodology we introduce establishes the groundwork for a global, data-driven early warning system designed to anticipate and detect food insecurity.

Anthropometric Profile and Position-Specific Changes in Segmental Body Composition of Professional Football Players Throughout a Training Period.

Body and anthropometric profiles of football players vary depending on the physiological and technical demands imposed by different positions. The aim of this study was to evaluate the body composition of professional soccer players in relation to their position on the field during a training macrocycle. The Direct Segmental Multi-Frequency Bioelectrical Impedance Analysis method was used to analyze 58 players at six key moments of the macrocycle. The results show that body profiles are adjusted to the specific demands of each position. Midfielders showed the lowest muscle mass, while defenders showed many notable changes. In general, as the season progressed, all field players experienced an increase in trunk body fat. Fat and lean mass values of goalkeepers differed greatly from the rest. The greatest variations in body composition were observed during pre-season and transition in relation to variations in training load and competitive intensity. The results suggest that the phase of the macrocycle has a greater influence on these variations, although the physical characteristics of each position are relevant. Understanding these dynamics allows for the design of more personalized and efficient training programs to optimize the performance of footballers according to their roles and each stage of the season.

Efficient training of energy-based models using Jarzynski equality**This article is an updated version of: Carbone D, Hua M, Coste S and Vanden-Eijnden E 2023 Efficient training of energy-based models using Jarzynski equality Advances in Neural Information Processing Systems vol 36, ed A Oh, T Naumann, A Globerson, K Saenko, M Hardt and S Levine (Curran Associates, Inc.) pp 52583–614.

Abstract Energy-based models (EBMs) are generative models inspired by statistical physics with a wide range of applications in unsupervised learning. Their performance is well measured by the cross-entropy (CE) of the model distribution relative to the data distribution. Using the CE as the objective for training is, however, challenging because the computation of its gradient with respect to the model parameters requires sampling of the model distribution. Here, we show how the results for nonequilibrium thermodynamics based on the Jarzynski equality together with tools from sequential Monte Carlo sampling can be used to perform this computation efficiently and avoid the uncontrolled approximations made using the standard contrastive divergence algorithm. Specifically, we introduce a modification of the unadjusted Langevin algorithm (ULA), in which each walker acquires a weight that enables the estimation of the gradient of the CE at any step during gradient descent, thereby bypassing sampling biases induced by slow mixing of the ULA. We illustrate these results with numerical experiments on Gaussian mixture distributions as well as the MNIST and CIFAR-10 datasets. We show that the proposed approach outperforms methods based on the contrastive divergence algorithm in all the considered situations.

Mamba-in-Mamba: Centralized Mamba-Cross-Scan in Tokenized Mamba Model for Hyperspectral image classification

Hyperspectral image (HSI) classification plays a crucial role in remote sensing (RS) applications, enabling the precise identification of materials and land cover based on spectral information. This supports tasks such as agricultural management and urban planning. While sequential neural models like Recurrent Neural Networks (RNNs) and Transformers have been adapted for this task, they present limitations: RNNs struggle with feature aggregation and are sensitive to noise from interfering pixels, whereas Transformers require extensive computational resources and tend to underperform when HSI datasets contain limited or unbalanced training samples. To address these challenges, Mamba architectures have emerged, offering a balance between RNNs and Transformers by leveraging lightweight, parallel scanning capabilities. Although models like Vision Mamba (ViM) and Visual Mamba (VMamba) have demonstrated improvements in visual tasks, their application to HSI classification remains underexplored, particularly in handling land-cover semantic tokens and multi-scale feature aggregation for patch-wise classifiers. In response, this study introduces the Mamba-in-Mamba (MiM) architecture for HSI classification, marking a pioneering effort in this domain. The MiM model features: (1) a novel centralized Mamba-Cross-Scan (MCS) mechanism for efficient image-to-sequence data transformation; (2) a Tokenized Mamba (T-Mamba) encoder that incorporates a Gaussian Decay Mask (GDM), Semantic Token Learner (STL), and Semantic Token Fuser (STF) for enhanced feature generation; and (3) a Weighted MCS Fusion (WMF) module with a Multi-Scale Loss Design for improved training efficiency. Experimental results on four public HSI datasets—Indian Pines, Pavia University, Houston2013, and WHU-Hi-Honghu—demonstrate that our method achieves an overall accuracy improvement of up to 3.3%, 2.7%, 1.5%, and 2.3% over state-of-the-art approaches (i.e., SSFTT, MAEST, etc.) under both fixed and disjoint training-testing settings.

Securing constrained IoT systems: A lightweight machine learning approach for anomaly detection and prevention

With the advent of advanced technological developments such as IoT, edge, and fog computing, cyber attacks have become increasingly sophisticated. IoT networks facilitate collaborative and intelligent tasks across various domains, including Industry 4.0, digital healthcare, and home automation. However, the proliferation of IoT devices has raised concerns about severe attacks, particularly those targeting resource constraints such as energy and memory. In response to these challenges, Tiny Machine Learning (TinyML) has emerged as a new research area, focusing on machine learning techniques tailored for embedded and IoT systems. This study proposes an ML detection mechanism designed to categorize and detect resource-constrained attacks in IoT devices. We consider IoT devices to be integral components within the continuum of edge and cloud computing, leveraging EdgeML and CloudML for detection purposes. Our paper conducts a comparative analysis of ML models, with a specific focus on energy consumption and memory usage in IoT applications. We compare various ML methodologies, including cloud-based, edge-based, and device-based strategies for both training and detection. The evaluation encompasses the application of these ML techniques to petite IoT devices, utilizing TinyML, as well as cloud and edge devices. Our findings reveal that the Decision Tree algorithm deployed on smart devices surpasses other approaches in terms of training efficiency, resource utilization, and the ability to detect resource-constrained attacks on IoT devices. We demonstrate a high level of accuracy, exceeding 96.9%, across all presented ML models in detecting resource constraint attacks within IoT systems. In summary, this research serves as a guide for implementing effective security measures in the dynamic landscape of IoT and associated technologies.