Model-based Training Research Articles

Mutual-guided scale-aggregation denoising network for seismic noise attenuation

The background noise contained in seismic records contaminate the effective reflection waves and impact the subsequent processes, such as inversion and migration. The properties of seismic noises, such as non-Gaussianity and non-linearity, will be even more complex in challenging exploration environments. Deep-learning techniques are effective in suppressing complex seismic noises and outperform conventional denoising algorithms. Nonetheless, most deep learning networks are designed to extract the features of input data in single-scale only, which leads to inadequate performance when dealing with complicated seismic data. To enhance the denoising capability for seismic noises of deep learning, a novel mutual-guided scale-aggregation denoising network (MSD-Net) is designed to suppress seismic noises by utilizing the multi-scale features of input data. Specifically, the MSD-Net achieves functions including multi-scale feature extraction, fusion, and guidance through information interaction between different scales. Spatial aggregation attention is used in MSD-Net to enhance relevant features, which improves the separation of effective reflection waves and noises further. Additionally, a model-based training data generation strategy is devised to ensure the efficiency of learning and the denoising capability of MSD-Net. Compared to conventional denoising algorithms and typical deep learning networks, MSD-Net shows powerful result in suppressing complex seismic noises and generalization.

YOLOv8-BYTE: Ship tracking algorithm using short-time sequence SAR images for disaster response leveraging GeoAI

Ship tracking technology is crucial for emergency rescue in the event of a disaster. Quickly identifying the position and status of vessels is vital for rescue teams to be able to deploy efficiently in disaster areas. When responding to emergencies or natural disasters, ship tracking technology plays a critical role in supporting emergency rescue operations and resource allocation, improving the overall resilience of the maritime transportation system. However, the research on multi-object tracking (MOT) algorithms has primarily focused on optical image datasets. In contrast, image data from synthetic aperture radar (SAR) presents unique challenges, such as defocus interference, a high false alarm rate, and a lack of prior samples. To overcome these particular challenges, we propose a robust MOT algorithm developed for SAR images to achieve effective multi-vessel tracking under difficult imaging conditions. In particular, we optimize the YOLOv8 detection network by introducing a diffusion model-based training method for data augmentation. This method improves the robustness of the network to scaling, rotational and translational deformations. Moreover, an enhanced swin transformer is proposed as a feature extraction network, which strengthens the representation capability of the detection network. Furthermore, the state parameters within the KF technique are enhanced by directly capturing the details of the height and width of the tracking rectangle box. This refinement of the ByteTrack algorithm aims to achieve a more precise and accurate fit of the tracking rectangle to the ship, further improving the overall tracking performance. The experimental results from the ship detection and multiple objects tracking datasets show the impressive performance of the proposed model. With a precision of 97.60%, a recall of 96.36%, and an average precision of 96.72%, the model achieves exceptional detection accuracy with an 18% reduction in model parameters. Furthermore, significant improvements can be observed in key tracking metrics such as HOTA, MOTA and IDF1, with improvements of 4.8%, 8.5% and 6.8% respectively compared to the baseline algorithm, alongside a remarkable 37.5% reduction in IDS. It is noteworthy that the tracker works in real time, achieving an average analysis speed of 47 frames per second. The proposed MOT algorithm achieves state-of-art tracking performance on a SAR image dataset with short time sequences. Therefore, the proposed approach is a compelling solution for ship tracking in SAR imagery.

Addie Model-Based Training Management As An Effort To Improve Employee Competence (Case Study At Ramayana Department Store Salatiga)

The purpose of this study is to investigate how Ramayana Department Store Salatiga has implemented training management using the five stages of the ADDIE model: analysis, design, development, implementation, and evaluation. An exploratory qualitative method was employed in this investigation. Interviews with many Ramayana Department Store resource people were done. Five individuals served as the study's resources: the head office, the store manager, the HR supervisor, the consignment team, and the employees. Based on the study's findings, Ramayana has implemented training management using ADDIE in line with many stages, including analysis, planning, development, implementation, and assessment. In order to ensure that this training supports Ramayana's vision and goal, it is necessary to design a few stages, namely the implementation and assessment stages. Ramayana has performed training to the fullest extent possible at each level.

Trainee Experience in Levels of Difficulties for Flaps in Plastic Surgery

The competency-based medical education (CBME) approach has introduced medical curricula across the globe, designed to provide flexibility, Responsibility and learner-centredness for medical learners. While traditional surgical skills training in most places has relied on "see one, do one, teach one model", simulation model-based training has been shown to improve surgical interns skills. Safely performing local and regional flaps is often a challenge for surgical residents. Therefore, the aim of this study was to assess the difficulty classification of flaps and establish equivalence between them at each level. A questionnaire and an assessment of surgical skills were sent to all plastic surgery residents and young specialists in Morocco analyzing the theoretical and procedural knowledge of flaps. Skills are assessed using a modified version of an objective structured assessment of the Technical Skills Score.

Multi-objective optimization model-based training program development for athletes in college physical education

Abstract Digital technology has been used in a large number of applications in all aspects of life and various fields. In the field of sports, digital training has been increasingly emphasized by frontline coaches, athletes, and sports researchers. In this paper, by building an objective optimization model, combining particle swarm optimization, and optimizing the computation process with the Kriging agent model, we formulate the most suitable training plan for college athletes and improve the scientific level of college athletes’ training. The multi-objective optimization training plan can accurately enhance the athletes’ physical fitness, physical quality, and sports skills, and the IRM and maximal power of bench press and half squat showed an increasing trend after the training. The body composition (body fat) and longitudinal jump (lower limb explosive power) improved to a certain extent before and after the training. There were significant differences in the sports skill evaluation of the 50-meter run and the three-stage continuous frog jump; the p-value was less than 0.05. The p-values were all less than 0.05. The scientific and practicality of the training program for college athletes based on the multi-objective optimization model is confirmed when the value falls below 0.05.

The introduction of surgical simulation on three-dimensional-printed models in the cardiac surgery curriculum: an experimental project.

Training in congenital cardiac surgery has become more and more difficult because of the reduced opportunities for trainees in the operating room and the high patient anatomical variability. The aim of this study was to perform a pilot evaluation of surgical simulation on a simple 3D-printed heart model in training of young surgeons and its potential inclusion in the curriculum of residency programs. A group of 11 residents performed a surgical correction of aortic coarctation using a 3D-printed surgical model. After teaching the surgical procedure, a simulation was performed twice, at different times, and was evaluated quantitatively and qualitatively by a senior surgeon. A 3D model-based training program was then developed and incorporated into our cardiac surgery training program. A significant improvement in surgical technique was observed between the first and second surgical simulations: median of 65% [interquartile range (IQR) = 61-70%] vs. 83% (IQR = 82-91%, P < 0.001). The median time required to run the simulation was significantly shorter during the second simulation: 39 min (IQR = 33-40) vs. 45 min (IQR = 37-48; P = 0.02). Regarding the simulation program, a basic and an advanced program were developed, including a total of 40 different simulated procedures divided into 12 sessions. Surgical simulation using 3D-printing technology can be an extremely valuable tool to improve surgical training in congenital heart disease. Our pilot study can represent the first step towards the creation of an integrated training system on 3D-printed models of congenital and acquired heart diseases in other Italian residency programs.

Objective structured assessment of medical students’ technical skills in second-degree perineal laceration repair with sponge model-based training

PurposeIn this cohort study, we used a sponge simulator to train students in second-degree perineal laceration repair. We examined whether the training course improved the students’ skills, as measured with an objective structured assessment of technical skills (OSATS) and by a senior physician. We also examined the correlation between these ratings to assess the validity of OSATS application in this context.MethodsBetween April and July 2022, 40 medical students took part in gynecological/obstetrics training that included a lecture about perineal trauma and the viewing of a video that demonstrated second-degree perineal laceration repair using a sponge model. They then underwent initial evaluation by a senior physician and OSATS application, yielding two independent scores. After training with the sponge model, a second evaluation was performed. The OSATS assessed practical skills (8 items) and suture results (2 items). The senior physician assigned ratings on a five-point ordinal scale ranging from 1 (excellent) to 5 (poor).ResultsTraining with the sponge simulator significantly increased students’ OSATS (practical skills, p < 0.001; suture results, p < 0.05) and senior physician (p < 0.001) ratings. The OSATS and senior physician ratings correlated strongly (Spearman’s r: first assessment, – 0.72; second assessment, – 0.74; p < 0.01).ConclusionThe sponge-based training improves students’ skills for the repair of a second-degree perineal laceration. The OSATS for the sponge model might be a valid option to examine medical students in an obstetrical course.

Detection of Unknown DDoS Attack Using Convolutional Neural Networks Featuring Geometrical Metric

DDoS attacks remain a persistent cybersecurity threat, blocking services to legitimate users and causing significant damage to reputation, finances, and potential customers. For the detection of DDoS attacks, machine learning techniques such as supervised learning have been extensively employed, but their effectiveness declines when the framework confronts patterns exterior to the dataset. In addition, DDoS attack schemes continue to improve, rendering conventional data model-based training ineffectual. We have developed a novelty open-set recognition framework for DDoS attack detection to overcome the challenges of traditional methods. Our framework is built on a Convolutional Neural Network (CNN) construction featuring geometrical metric (CNN-Geo), which utilizes deep learning techniques to enhance accuracy. In addition, we have integrated an incremental learning module that can efficiently incorporate novel unknown traffic identified by telecommunication experts through the monitoring process. This unique approach provides an effective solution for identifying and alleviating DDoS. The module continuously improves the model’s performance by incorporating new knowledge and adapting to new attack patterns. The proposed model can detect unknown DDoS attacks with a detection rate of over 99% on conventional attacks from CICIDS2017. The model’s accuracy is further enhanced by 99.8% toward unknown attacks with the open datasets CICDDoS2019.

A Bayesian-based knowledge tracing model for improving safety training outcomes in construction: An adaptive learning framework

Safety training plays a pivotal role in effectively reducing unsafe behaviors in the construction industry. Despite the numerous effort to improve safety training, existing studies lacked consideration of trainees' different learning characters, nor did they adapt the suitable training materials to trainees' mastery of safety knowledge. Against this contextual backdrop, this research proposes a Bayesian-based Knowledge Tracing (BKT) model to recommend personalized training materials according to trainees's learning progress. The proposed BKT model tracks and predicts trainees' performance with their cognitive characters, abilities, and historical training records. It also adjusts trainee's probability of mastering safety knowledge concepts for determining future training sessions. In this way, the safety training is adaptive and personalized, thus more effective. An exploration study is conducted in the laboratory environment to validate the effectiveness and feasibility of the proposed BKT model. The research results showed that 83.33% of the respondents felt gained safety knowledge after the training, and 66.67% of the respondents affirmed the effectiveness of the BKT model-based training system in improving the effectiveness of training. The results demonstrated that the proposed model performed well at educating safety knowledge and reducing unsafe behavior.

Animal model-based simulation training for three emergent and urgent operations of penetrating thoracic injuries

PurposeTo develop animal models of penetrating thoracic injuries and to observe the effects of the animal model-based training on improving the trainees’ performance for emergent and urgent thoracic surgeries. MethodsWith a homemade machine, animal models of lung injuries and penetrating heart injuries were produced in porcine and used for training of chest tube drainage, urgent sternotomy, and emergent thoracotomy. Coefficient of variation of abbreviated injury scale and blood loss was calculated to judge the reproducibility of animal models. Five operation teams from basic-level hospitals (group A) and five operation teams from level III hospitals (group B) were included to be trained and tested. Testing standards for the operations were established after thorough literature review, and expert questionnaires were employed to evaluate the scientificity and feasibility of the testing standards. Tests were carried out after the training. Pre- and post-training performances were compared. Post-training survey using 7-point Likert scale was taken to evaluate the feelings of the trainees to these training approaches. ResultsAnimal models of the three kinds of penetrating chest injuries were successfully established and the coefficient of variation of abbreviated injury scale and blood loss were all less than 25%. After literature review, testing standards were established, and expert questionnaire results showed that the scientific score was 7.30 ± 1.49, and the feasibility score was 7.50 ± 0.89. Post-training performance was significantly higher in both group A and group B than pre-training performance. Post-training survey showed that all the trainees felt confident in applying the operations and were generally agreed that the training procedure were very helpful in improving operation skills for thoracic penetrating injury. ConclusionsAnimal model-based simulation training established in the current study could improve the trainees’ performance for emergent and urgent thoracic surgeries, especially of the surgical teams from basic-level hospitals.

Model-based and model-free decision making in major depressive disorder after performing behavioral training

IntroductionIn major depressive disorder (MDD), reward-based decision-making (DM) is frequently impaired: e.g. patients don’t engage in pleasant activities as much as healthy subjects. Put differently, previous and expected future rewards have less reinforcing effects on DM. This study investigated two experimentally well-observable reward-based DM modes, namely model-based (based on cognitive models of the environment) and model-free (based on previous experience) DM.ObjectivesWe hypothesized that model-based training can improve reward-based DM in patients with MDD. Answers to these questions could enhance the development of cognitive-behavioral therapeutic interventions.Methods27 patients with MDD were recruited and assessed with psychometry. All patients performed the „two-step Markov decision-task“ (Daw, 2011), which allows the simultaneous investigation of model-based and model-free DM via computational modelling. All subjects performed the task 4 times: at the beginning and at the end of 2 assessment days (session-interval: 4 days). Subjects were randomly allocated to an intervention group, which performed model-based training, and a control group, which performed model-free training. The main outcomes of training effect were the influence of model-based reward expectations on decisions (quantified by computational modelling parameters) and overall monetary reward-success.ResultsIn all patients, the influence of model-based reward expectations on decisions increased after training. However, there was no significant effect of group allocation. Furthermore, patients in the intervention group did not achieve significantly higher overall monetary reward.ConclusionsResults suggest that in MDD, the influence of model-based reward expectations on decisions can be improved regardless of specific training type. Future studies should investigate the effects on everyday functioning.DisclosureNo significant relationships.

DisCOV: Distributed COVID-19 Detection on X-Ray Images With Edge-Cloud Collaboration

Currently, the world is experiencing the rapid spread of Coronavirus Disease 2019 (COVID-19). Since the epidemic continues to take a devastating impact on the society, economy, and healthcare, the real-time detection of COVID-19 is essential for fast and cost-effective diagnosis services. Fortunately, deep learning (DL), as a promising technology, enables the COVID-19 diagnosis services on chest X-ray (CXR) images. The training task of DL model is generally implemented at the centralized cloud. However, due to the geo-distributed data sources and the transmission of large amounts of raw data to the centralized cloud, the transmission latency becomes a bottleneck of the COVID-19 diagnosis model training. In this paper, we propose a <u>Dis</u>tributed <u>COV</u>ID-19 detection model training method on CXR images with edge-cloud collaboration, named DisCOV. Specifically, to improve the training efficiency and guarantee the model accuracy, a distributed lightweight model-based training algorithm is designed with the cooperation of edge computing and cloud computing. In addition, a resource allocation algorithm is developed during the training to jointly minimize the time cost and energy consumption. Extensive experiments based on real-world CXR image datasets demonstrate that DisCOV is better performed and more promising than the existing baselines.

An Innovative Skin Simulation Model to Augment Competency-based Training in Facial Plastic Surgery.

Introduction Competency-based medical education (CBME) approach in the medical curriculum has been introduced globally with the goal of providing flexibility, accountability, and learner-centeredness among medical learners. Traditional surgical skill training in most places has relied on “see one, do one, teach one model,” while simulation model-based training has been shown to improve competencies in surgical trainees. We wanted to assess the usefulness of a hydrophilic barrier adhesive foam wound dressing as a novel skin simulation model for learning biomechanics and practice of cutaneous flaps among plastic surgical resident trainees at our institute. Materials and Methods An absorbent, soft polyurethane foam pad located centrally upon a larger polyurethane membrane, coated with a hydrocolloid adhesive, forming an island dressing, was used as a simulation model for this study. It was obtained from the hospital store either after or nearing their expiry dates of clinical use. Plastic surgery residents in different years of training were invited to participate in a simulation workshop, using this novel model, and give their feedback. Results Seventeen residents in different plastic surgery training levels participated in the workshop and gave their feedback on the skin flap simulation model. The simulation model received extremely high (100%) scores on two parameters, namely, utility for flap and suture practice and high scores (88%–94%) for texture, ability to mark, and improving confidence among trainees. Conclusions Adhesive bilayer polyurethane foam can be used as a novel cutaneous skin flap simulation model for understanding the biomechanics of skin flaps and cutaneous flap practice.

Fruit quality and defect image classification with conditional GAN data augmentation

Contemporary Artificial Intelligence technologies allow for the employment of Computer Vision to discern good crops from bad, providing a step in the pipeline of selecting healthy fruit from undesirable fruit, such as those which are mouldy or damaged. State-of-the-art works in the field report high accuracy results on small datasets (<1000 images), which are not representative of the population regarding real-world usage. The goals of this study are to further enable real-world usage by improving generalisation with data augmentation as well as to reduce overfitting and energy usage through model pruning. In this work, we suggest a machine learning pipeline that combines the ideas of fine-tuning, transfer learning, and generative model-based training data augmentation towards improving fruit quality image classification. A linear network topology search is performed to tune a VGG16 lemon quality classification model using a publicly-available dataset of 2690 images. We find that appending a 4096 neuron fully connected layer to the convolutional layers leads to an image classification accuracy of 83.77%. We then train a Conditional Generative Adversarial Network on the training data for 2000 epochs, and it learns to generate relatively realistic images. Grad-CAM analysis of the model trained on real photographs shows that the synthetic images can exhibit classifiable characteristics such as shape, mould, and gangrene. A higher image classification accuracy of 88.75% is then attained by augmenting the training with synthetic images, arguing that Conditional Generative Adversarial Networks have the ability to produce new data to alleviate issues of data scarcity. Finally, model pruning is performed via polynomial decay, where we find that the Conditional GAN-augmented classification network can retain 81.16% classification accuracy when compressed to 50% of its original size.

Efficacy of Health Belief Model-Based Training in Changing the Beliefs about Substance use

Background: Substance use problems are associated with capitalizes health retro gradation, weakness and death due to impacts and exceed.Aims of the study: The purpose of this study was to identify the efficacy of health belief model based training in changing the beliefs about substance use among university students at Mosul City.Methodology: This study is based on experimental design by using the random controlled trial approach is conducted. A probability (simple random sample) of (N=80) undergraduate student. Data were gathered in 1st of October / 2019 till 12 of May / 2020 using a questionnaire method. The instrument consist of two parts; part I, involved, to describe the student's socio-demographic characteristic such as (age, gender, grade, college ,socioeconomic status) the part II, involved the using scale to measure students' beliefs towards of substance use.Results: demonstrated that the results of study the mean ± SD age of the study group and control group was (23.37 ± 2.09) and (23.70 ± 2.04) respectively. In addition, the same table demonstrated that majority of participant were normal body weight, the overall mean (± SD) of the body mass index were (21.23 ± 2.53), and the Mean scores and the Standard Deviations for the variables under the study were changed among participant's over times.Conclusion: This study concluded that designing an HBM-based study could affect students' understanding and their behaviors in the field of substance abuse.Recommendations: This study recommended there is a need to conduct future studies based on the Health Belief Model on large number of the Iraqi population with the goal of changing People's behavior about addiction.

Predictions-on-chip: model-based training and automated deployment of machine learning models at runtime

The design of gas turbines is a challenging area of cyber-physical systems where complex model-based simulations across multiple disciplines (e.g., performance, aerothermal) drive the design process. As a result, a continuously increasing amount of data is derived during system design. Finding new insights in such data by exploiting various machine learning (ML) techniques is a promising industrial trend since better predictions based on real data result in substantial product quality improvements and cost reduction. This paper presents a method that generates data from multi-paradigm simulation tools, develops and trains ML models for prediction, and deploys such prediction models into an active control system operating at runtime with limited computational power. We explore the replacement of existing traditional prediction modules with ML counterparts with different architectures. We validate the effectiveness of various ML models in the context of three (real) gas turbine bearings using over 150,000 data points for training, validation, and testing. We introduce code generation techniques for automated deployment of neural network models to industrial off-the-shelf programmable logic controllers.

A proposed model-based training program to develop the teaching performance of science teachers in Islamic high schools

The aim of the current research is to know the effect of a proposed program for professional development based on modules for science teachers in Islamic high schools in Iraq, and to achieve the research objectives, a list of the teaching performances necessary to develop the teaching performance of science teachers was built, and through those training needs a proposed program for the development of these teaching performances was built based on (Modules), and an achievement test was built to measure the cognitive side and a note card to measure the skill side.&#x0D; The achievement test and the observation card were applied beforehand to the research group, then the proposed program was presented to them, then the achievement test and the observation card were presented to them afterward, and by measuring the differences between the results of the pre and post application of the achievement test and the observation card, it was found that the proposed program is effective in developing the knowledge and skills side of science teachers The research sample, and in light of the research results, the researcher presented a set of recommendations and suggestions.

Effect of Model-Based Training by Nurses on the Promotion of Sleep Patterns among Students in the Middle School

Effect of Model-Based Training by Nurses on the Promotion of Sleep Patterns among Students in the Middle School

Rational Model-Based Training Retain the Health Cadres’ Knowledge, Attitudes and Practices on Stroke Issue

Stroke is a medical emergency that the patients of it ideally should have received stroke management ?3 hours since the attack. Emergency situations can occur at anytime, thus the health cadres participation in minimizing pre-hospital delay through FAST method is needed. In order to make the cadres have knowledge, positive attitudes, and good practices about FAST, the rational model-based training was conducted. The aim of this study was to analyze the effectiveness of rational model-based training on the health cadres’ changes on knowledge, attitudes, and practices in the stroke symptoms detection using FAST. This study used quasi experimental research design through pretest-posttest control group design. 50 respondents were divided into intervention and control groups. Data analysis used independent sample t-test and mann whitney test. The results show that there were differences in attitude and practices changes between the intervention and the control groups after training, with p values 0.009 and 0.000. There was no any difference in the retention of knowledge, attitudes, and practices in both groups, the p value was 0.849; 0.626;0.456.

The Value of a Synthetic Model-based Training Lab to Increase Proficiency with Endoscopic Approaches to the Spine.

Introduction: The learning curve associated with endoscopic approaches to the spine is well established. In this study, we present our endoscopic training methodology and discuss the concepts and rationale of laboratory training as it relates to improving comfort and skill with endoscopic techniques.Materials and Methods: A three-part endoscopic training laboratory for neurosurgical trainees and attendings was organized at the University of Miami, which included a lecture, instrumentation demonstration, and both synthetic model and cadaveric practice sessions. Participants completed pre- and post-lab surveys gauging their comfort and competency in the transforaminal approach to the lumbar spine.Results: There were a total of 22 participants, with eight completing the pre-lab survey and 10 completing the post-lab survey. Sixteen participants engaged in the lab practical, with six of these participants performing the transforaminal approach on both the model and the cadaver. An increase in comfort level was demonstrated on the post-lab survey (5.9/10) for the transforaminal approach as compared to the pre-lab survey (2.6/10). Additionally, participants found the training model to be an effective teaching aid for the transforaminal technique (8.8/10).Conclusions: We believe that our study demonstrates the utility of simulated model-based training for gaining comfort and proficiency with endoscopic approaches to the spine and introduces a safe, cost-effective method of educating practitioners on novel endoscopic approaches.