Early Stages Of Training Research Articles

CalcGraph: taming the high costs of deep learning using models

Models based on differential programming, like deep neural networks, are well established in research and able to outperform manually coded counterparts in many applications. Today, there is a rising interest to introduce this flexible modeling to solve real-world problems. A major challenge when moving from research to application is the strict constraints on computational resources (memory and time). It is difficult to determine and contain the resource requirements of differential models, especially during the early training and hyperparameter exploration stages. In this article, we address this challenge by introducing CalcGraph, a model abstraction of differentiable programming layers. CalcGraph allows to model the computational resources that should be used and then CalcGraph’s model interpreter can automatically schedule the execution respecting the specifications made. We propose a novel way to efficiently switch models from storage to preallocated memory zones and vice versa to maximize the number of model executions given the available resources. We demonstrate the efficiency of our approach by showing that it consumes less resources than state-of-the-art frameworks like TensorFlow and PyTorch for single-model and multi-model execution.

A questionnaire study on the knowledge, attitudes, and practices of fluid replacement and urination among Chinese elite athletes.

ObjectiveTo evaluate the knowledge, attitudes and practices (KAP) of Chinese elite athletes about fluid replacement and urination.MethodsA cross-section study was carried out among Chinese national and national youth teams from March to April 2020, using a pretested questionnaire. The 42-questions questionnaire was designed to assess the KAP regarding fluid replacement and urination. The questionnaire included knowledge of fluid replacement (KFR), attitudes of fluid replacement (AFR), knowledge of urination (KU), and attitudes of urination (AU), which were awarded 20 scoring points. Descriptive statistics, independent samples t-tests, one-way ANOVA, Pearson’s correlation analysis, Multiple linear stepwise regression and Chi-square test were performed.ResultsA total of 779 valid questionnaires were collected and the effective rate is 98.4%. We finally conducted an assessment of 646 questionnaires of elite athletes. The mean score for KFR, AFR, KU, and AU was 2.8±1.3, 2.3±0.6, 3.0±1.5, and 2.1±0.8, respectively, with higher scores indicating positive hydration knowledge and attitudes. KFR and AFR scores of winter sports athletes were higher than those of summer sports athletes(P<0.05). Athletes who had lower athletic grades and training years had a worse KFR(P<0.05). Only 31.0% athletes knew that rehydration should be carried out before, during, and after training, which was scarcer among women, lower-athletic grades athletes, or athletes with lower training years (P<0.05). Male athletes had a worse KU but a better AU than female athletes(P<0.05). And athletes who were international-class athletic grades had the highest KU scores(P<0.05). The athletic grades and sport events were the main factors influencing the total scores of knowledge and attitudes (P<0.05, 95% CI -0.789–-0.168,95% CI 0.025–1.040). Most of athletes tend to get hydration knowledge from internet. In practices, thirst is the main reason for rehydration (77.9%). The percentages of athletes with normal urine color (42.0%), frequency (75.0%,) and volume (20.0%) were low.ConclusionsThese findings indicate that Chinese elite athletes did not have sufficient KAP on fluid replacement and urination, more marked in the individuals who were summer sport events, the lower athletic grades and in lower training years. It is recommended that education should be provided in the early stages of professional training for athletes.

Augmented Hill-Climb increases reinforcement learning efficiency for language-based de novo molecule generation

A plethora of AI-based techniques now exists to conduct de novo molecule generation that can devise molecules conditioned towards a particular endpoint in the context of drug design. One popular approach is using reinforcement learning to update a recurrent neural network or language-based de novo molecule generator. However, reinforcement learning can be inefficient, sometimes requiring up to 105 molecules to be sampled to optimize more complex objectives, which poses a limitation when using computationally expensive scoring functions like docking or computer-aided synthesis planning models. In this work, we propose a reinforcement learning strategy called Augmented Hill-Climb based on a simple, hypothesis-driven hybrid between REINVENT and Hill-Climb that improves sample-efficiency by addressing the limitations of both currently used strategies. We compare its ability to optimize several docking tasks with REINVENT and benchmark this strategy against other commonly used reinforcement learning strategies including REINFORCE, REINVENT (version 1 and 2), Hill-Climb and best agent reminder. We find that optimization ability is improved ~ 1.5-fold and sample-efficiency is improved ~ 45-fold compared to REINVENT while still delivering appealing chemistry as output. Diversity filters were used, and their parameters were tuned to overcome observed failure modes that take advantage of certain diversity filter configurations. We find that Augmented Hill-Climb outperforms the other reinforcement learning strategies used on six tasks, especially in the early stages of training or for more difficult objectives. Lastly, we show improved performance not only on recurrent neural networks but also on a reinforcement learning stabilized transformer architecture. Overall, we show that Augmented Hill-Climb improves sample-efficiency for language-based de novo molecule generation conditioning via reinforcement learning, compared to the current state-of-the-art. This makes more computationally expensive scoring functions, such as docking, more accessible on a relevant timescale.

Robust Adaptive Ensemble Adversary Reinforcement Learning

Reinforcement learning needs to learn policies through trial and error. The unstable policies in the early stage of training make it expensive (and time-consuming) to train directly in the real environment, which may cause disastrous consequences. The popular solution is to use the simulator to train the policy and deploy it in a real environment. However, the modeling error and external disturbance between the simulation and the real environment may fail the physical deployment, resulting in the sim2real transfer problem. In this letter, we propose a novel robust adversarial reinforcement learning framework, which uses the ensemble training of multi-adversarial agents that can adaptively adjust adversaries' strength to enhance RL policy's robustness. More specifically, we take the accumulative reward as feedback and construct a PID controller to adjust the adversary's output magnitude to perform the adversarial training well. Experiments in the simulated and the real environment show that our algorithm improves the generalization ability of the policy for the modeling error and the uncertain disturbance simultaneously, outperforming the next best prior methods across all domains. The algorithm was further proven to be effective in a sim2real transfer task through the load experiment of a real racing drone, and the tracking performance is better than the PID-based flight controller.

A Fault Detection Method Based on an Oil Temperature Forecasting Model Using an Improved Deep Deterministic Policy Gradient Algorithm in the Helicopter Gearbox

The main gearbox is very important for the operation safety of helicopters, and the oil temperature reflects the health degree of the gearbox; therefore establishing an accurate oil temperature forecasting model is an important step for reliable fault detection. Firstly, in order to achieve accurate gearbox oil temperature forecasting, an improved deep deterministic policy gradient algorithm with a CNN–LSTM basic learner is proposed, which can excavate the complex relationship between oil temperature and working condition. Secondly, a reward incentive function is designed to accelerate the training time costs and to stabilize the model. Further, a variable variance exploration strategy is proposed to enable the agents of the model to fully explore the state space in the early training stage and to gradually converge in the training later stage. Thirdly, a multi-critics network structure is adopted to solve the problem of inaccurate Q-value estimation, which is the key to improving the prediction accuracy of the model. Finally, KDE is introduced to determine the fault threshold to judge whether the residual error is abnormal after EWMA processing. The experimental results show that the proposed model achieves higher prediction accuracy and shorter fault detection time costs.

Predicting Performance at Graduation From Early ACGME Milestone Ratings: Longitudinal Learning Analytics in Professionalism and Communication in Vascular Surgery

Program directors in surgical disciplines need more tools from the ACGME to help them use Milestone ratings to improve trainees' performance. This is especially true in competencies that are notoriously difficult to measure, such as professionalism (PROF) and interpersonal and communication skills (ICS). It is now widely understood that skills in these two areas have direct impact on patient care outcomes. This study investigated the potential for generating early predictors of final Milestone ratings within the PROF and ICS competency categories. This retrospective cohort study utilized Milestone ratings from all ACGME-accredited vascular surgery training programs, covering residents and fellows who completed training in June 2019. The outcome measure studied was the rate of achieving the recommended graduation target of Milestone Level 4 (possible range: 1-5), while the predictors were the Milestone ratings attained at earlier stages of training. Predictive probability values (PPVs) were calculated for each of the 3 PROF and two ICS sub-competencies to estimate the probability of trainees not reaching the recommended graduation target based on their previous Milestone ratings. All ACGME-accredited vascular surgery training programs within the United States. All trainees completing a 2 year vascular surgery fellowship (VSF) in June 2019 (n = 119) or a 5 year integrated vascular surgery residency (IVSR) in June 2019 (n = 52) were included in the analyses. The overall rate of failing to achieve the recommended graduation target across all PROF and ICS sub-competencies ranged from 7.7% to 21.8% of all trainees. For trainees with a Milestone rating at ≤ 2.5 with 1 year remaining in their training program, the predictive probability of not achieving the recommended graduation target ranged from 37.0% to 71.5% across sub-competencies, with the highest risks observed under PROF for "Administrative Tasks" (71.5%) and under ICS for "Communication with the Healthcare Team" (56.7%). As many as 1 in 4 vascular surgery trainees did not achieve the ACGME vascular surgery Milestones targets for graduation in at least one of the PROF and ICS sub-competencies. Biannual ACGME Milestone assessment ratings of PROF and ICS during early training can be used to predict achievement of competency targets at time of graduation. Early clues to problems in PROF and ICS enable programs to address potential deficits early in training to ensure competency in these essential non-technical skills prior to entering unsupervised practice.

Sitting comfort analysis and prediction for high-speed rail passengers based on statistical analysis and machine learning

Various comfort evaluation models have been extensively studied for improving driver seat and office seat designs for the purpose of improving seat comfort. High-speed rail (HSR) seat comfort is influenced by specific environments, activities, and postures, and targeted in-depth research on this topic is lacking. Existing evaluation cannot ensure the validity of data at the early stages of model training because of their direct reliance on subjective evaluation data, thereby limiting the improvement in the accuracy of prediction. In this study, we designed a sitting observation experiment and static sitting comfort experiment for the typical activities of HSR passengers. The sitting posture of HSR passengers under typical activities was classified for different comfort levels based on the factors influencing the sitting posture in terms of bone and muscle biomechanics. The correlations between the subjective discomfort in different sitting positions, objective change patterns, and body pressure distribution parameters were statistically analysed. The possibility of using sitting duration and sitting frequency as objective behavioural indicators of discomfort evaluation was indicated. Based on the verification of the reliability of the evaluation data, the classification performance of various machine learning algorithms was compared and analysed, and a sitting comfort prediction model was established based on the gradient boosting machine algorithm, with an accuracy of 89.5%. This study serves as an effective reference for improvement strategies for HSR passenger comfort and will inspire new ideas for exploring objective indicators for sitting comfort evaluation.

Exploration Strategy Improved DDPG for Lane Keeping Tasks in Autonomous Driving

We propose an Exploration Strategy Improved Deep Deterministic Policy Gradient algorithm called ESI-DDPG for lane keeping tasks in autonomous driving. The actor network in DDPG outputs a policy which is deterministic, so it is necessary to add noise to actions so that the autonomous vehicle can fully explore the environment and learn the optimal policy. However, the initial weight of exploration noise is large, which makes the autonomous vehicle carry out a lot of invalid exploration in the early training stage. Therefore, we combine the Stanley method to make a weighted correction to the exploration noise, so that the exploration of the vehicle tends to be in the right direction and the training efficiency can be improved. In addition, due to the good sample data obtained in the training process, the driving policy finally learned is better. We choose TORCS as the experiment platform and the results show that, compared with DDPG, TD3 and SAC, our proposed algorithm can learn the driving policy faster and the final policy has smaller trajectory error while driving.

Res-CDD-Net: A Network with Multi-Scale Attention and Optimized Decoding Path for Skin Lesion Segmentation

Melanoma is a lethal skin cancer. In its diagnosis, skin lesion segmentation plays a critical role. However, skin lesions exhibit a wide range of sizes, shapes, colors, and edges. This makes skin lesion segmentation a challenging task. In this paper, we propose an encoding–decoding network called Res-CDD-Net to address the aforementioned aspects related to skin lesion segmentation. First, we adopt ResNeXt50 pre-trained on the ImageNet dataset as the encoding path. This pre-trained ResNeXt50 can provide rich image features to the whole network to achieve higher segmentation accuracy. Second, a channel and spatial attention block (CSAB), which integrates both channel and spatial attention, and a multi-scale capture block (MSCB) are introduced between the encoding and decoding paths. The CSAB can highlight the lesion area and inhibit irrelevant objects. MSCB can extract multi-scale information to learn lesion areas of different sizes. Third, we upgrade the decoding path. Every 3 × 3 square convolution kernel in the decoding path is replaced by a diverse branch block (DBB), which not only promotes the feature restoration capability, but also improves the performance and robustness of the network. We evaluate the proposed network on three public skin lesion datasets, namely ISIC-2017, ISIC-2016, and PH2. The dice coefficient is 6.90% higher than that of U-Net, whereas the Jaccard index is 10.84% higher than that of U-Net (assessed on the ISIC-2017 dataset). The results show that Res-CDD-Net achieves outstanding performance, higher than the performance of most state-of-the-art networks. Last but not least, the training of the network is fast, and good results can be achieved in early stages of training.

Evaluating the significance of samples in deep learning-based transient stability assessment

Deep learning-based transient stability assessment has achieved big success in power system analyses. However, it is still unclear how much of the data is superfluous and which samples are important for training. In this work, we introduce the latest technique from the artificial intelligence community to evaluate the significance of the samples used in deep learning model for the transient stability assessment. From empirical experiments, it is found that nearly 50% of the low-significance samples can be pruned without affecting the testing performance at the early training stages, thus saving much computational time and effort. We also observe that the samples with the fault-clearing time close to the critical clearing time often have higher significance indexes, indicating that the decision boundary learned by the deep network is highly related to the transient stability boundary. This is intuitive, but to the best of our knowledge, this work is the first to analyze the connection from sample significance aspects. In addition, we combine the stability scores with the significance index to provide an auxiliary criterion for the degree of stability, indicating the distance between a sample and the stability boundary. The ultimate goal of the study is to create a tool to generate and evaluate some benchmark datasets for the power system transient stability assessment analysis, so that various algorithms can be tested in a unified and standard platform like computer vision or natural language-processing fields.

Multi action recognition using machine learning

One of the most exciting and useful computer vision research topics is automated human activity identification. The majority of existing research, as well as most traditional approaches and classic neural networks, disregard the In video sequences, the appearance and patterns of motion are important. Individuals are unableIn video sequences, the appearance and patterns of motion are important.. This paper outlines a system for detecting, recognising, and summarising diverse human actions.The multiple action detection method takes the silhouettes of human bodies and then uses motion detection and tracking to create a unique sequence for each one.Based on the Each of the recovered sequences is then separated into shots based on the similarity between each pair of frames.show the sequence's homogenous activity. The activity is identified by looking at thedata oriented gradient (HOG) histogram of the frames in each shot's Temporal Difference Map (TDMap). comparing generated HOG to previous HOGs in the early stages of training. Making use of the TDMap image and a proposed CNN model, we can additionally recognise the action. Action summarising is carried out for each person found.

TBR-NER: Research on COVID-19 Text Information Extraction Based on Joint Learning of Topic Recognition and Named Entity Recognition

There is a centralization of the core content in the text information of the new crown epidemic notification. This paper proposes a joint learning text information extraction method: TBR-NER (topic-based recognition named entity recognition) based on topic recognition and named entity recognition to predict the labeled risk areas and epidemic trajectory information in text information. Transfer learning and data augmentation are used to solve the problem of data scarcity caused by the initial local outbreak of the epidemic, and mutual understanding is achieved by topic self-labeling without introducing additional labeled data. Taking the epidemic cases in Hebei and Jilin provinces as examples, the reliability and effectiveness of the method are verified by five types of topic recognition and 15 types of entity information extraction. The experimental results show that, compared with the four existing NER methods, this method can achieve optimality faster through the mutual learning of each task at the early stage of training. The optimal accuracy in the independent test set can be improved by more than 20%, and the minimum loss value is significantly reduced. This also proves that the joint learning algorithm (TBR-NER) mentioned in this paper performs better in such tasks. The TBR-NER model has specific sociality and applicability and can help in epidemic prediction, prevention, and control.

Evaluation of a prerequisite course of histology implementation for Chinese students of eight-year medical programme: a mixed quantitative survey

BackgroundDue to insufficient basic medical knowledge and inappropriate learning strategies, students of 8-year medical programme encountered many obstacles in the initial stage of basic medicine learning. This study was to determine whether a prerequisite course can improve basic medicine learning performance and adjust learning strategies to adapt to basic medicine learning.MethodsA prerequisite course of histology was constructed by a two-round modified Delphi study. Seventy-four students of 8-year medical programme were subjected to two groups: the prerequisite course group (PC group) and non-prerequisite course group (NPC group). The PC group take part in the prerequisite course by student-centred blended learning approach but NPC group not. The PC and NPC group underwent requisite histology teaching activities after prerequisite course. Examination of the prerequisite course and requisite histology course were carried out. Effect of the prerequisite course was evaluated by an empirical method using a questionnaire-based approach.ResultsThe results of examinations showed students' scores of the PC group were significantly higher than those of students of NPC group in both prerequisite course and requisite histology examinations (P < 0.05). The results of questionnaires showed that students were satisfied with the prerequisite course, which was beneficial for uptake in medical knowledge, cultivation of clinical thinking and scientific research ability and adaptation in learning strategies (P < 0.01). Furthermore, our prerequisite course is conducive to subsequent courses learning, especially for pathology (P < 0.01).ConclusionOur prerequisite course could effectively supplement knowledge of basic medicine, improve clinical thinking and scientific research ability and adapt their learning strategies. These findings suggest that the prerequisite course is useful and should be introduced in medical curriculum reform at the early stages of basic medical training.

FEATURES OF TECHNICAL TRAINING OF YOUNG VOLLEYBALL PLAYERS

Introduction. One of the most important methodological conditions for the growth of technical skills in volleyball already at the early stages of long-term training is the relationship between the structure of movements and the level of development of physical qualities and abilities. Compliance of the level of physical fitness of an athlete with the parameters of his sports equipment possession is the most important condition for an effective method of technical training in volleyball. The article presents the results of a study aimed at the implementation of training programs for the technical training of 9-10 year old volleyball players.&#x0D; Purpose: to develop and experimentally test the effectiveness of exercise complexes for the development of physical and psychomotor abilities of young volleyball players (muscle strength of the hand and phalanges of the fingers, the speed of contraction of the muscles of the hands, the ability to differentiate muscle tension, etc.), which are of decisive importance in the process of technical training of athletes at the initial stage.&#x0D; Methodology. The following methods were used in the study: generalization and analysis of literary sources; pedagogical testing; biomedical; pedagogical observation; statistical data processing.&#x0D; Results. The conducted research confirmed the put forward hypothesis that the development of technical elements in volleyball already at the early stages of training is determined by the proper level of development of special physical qualities and abilities. At the end of the pedagogical experiment using the new methodology, volleyball players from the experimental group significantly outperformed the athletes from the control group in almost all parameters of technical readiness.&#x0D; Application area. The developed sets of exercises can be used by sports teachers when compiling work programs within the framework of basic and additional education at the Volleyball course.

Personalized On-Device E-Health Analytics With Decentralized Block Coordinate Descent.

Actuated by the growing attention to personal healthcare and the pandemic, the popularity of E-health is proliferating. Nowadays, enhancement on medical diagnosis via machine learning models has been highly effective in many aspects of e-health analytics. Nevertheless, in the classic cloud-based/centralized e-health paradigms, all the data will be centrally stored on the server to facilitate model training, which inevitably incurs privacy concerns and high time delay. Distributed solutions like Decentralized Stochastic Gradient Descent (D-SGD) are proposed to provide safe and timely diagnostic results based on personal devices. However, methods like D-SGD are subject to the gradient vanishing issue and usually proceed slowly at the early training stage, thereby impeding the effectiveness and efficiency of training. In addition, existing methods are prone to learning models that are biased towards users with dense data, compromising the fairness when providing E-health analytics for minority groups. In this paper, we propose a Decentralized Block Coordinate Descent (D-BCD) learning framework that can better optimize deep neural network-based models distributed on decentralized devices for E-health analytics. As a gradient-free optimization method, Block Coordinate Descent (BCD) mitigates the gradient vanishing issue and converges faster at the early stage compared with the conventional gradient-based optimization. To overcome the potential data scarcity issues for users' local data, we propose similarity-based model aggregation that allows each on-device model to leverage knowledge from similar neighbor models, so as to achieve both personalization and high accuracy for the learned models. Benchmarking experiments on three real-world datasets illustrate the effectiveness and practicality of our proposed D-BCD, where additional simulation study showcases the strong applicability of D-BCD in real-life E-health scenarios.

Use of Actors or Peers as Simulated Clients in Veterinary Communication Training.

Using simulated clients is an effective teaching method for training and assessing communication skills in veterinary education. The aim of this study is to evaluate the use of actors and peers in communication skills training in veterinary medicine. For this purpose, the subjective perception of the use of actors was assessed in a first study using a paper-based self-evaluation survey. In a second study, different groups of veterinary students who trained their communication skills with actors or peers were compared in an electronic Objective Structured Clinical Examination (eOSCE) assessment with regard to their outcomes of communication proficiency. All participants reported the actors to be helpful and supportive in learning communication skills. Above all, participants highly rated the achieved authenticity when using actors as well as feedback sessions. Regarding the comparison of actors and peers as teaching methods, no significant difference in the performance of veterinary students in an eOSCE was identified. Despite the lack of objective evidence, both methods may be considered valuable and accepted teaching tools. Training with peers gives students an opportunity to learn how to conduct structured history interviews and to understand pet owners' motives at an early stage of undergraduate veterinary training. Change of perspective is considered a positive training element. However, when portraying authentic and standardized emotions and reactions and giving formative feedback based on the pet owners' internal perspectives, actors are beneficial for training advanced veterinary students and graduates in difficult conversation topics.

Unsupervised English Intelligent Machine Translation in Wireless Network Environment

Researchers suggest unsupervised English machine translation to address the absence of parallel corpus in English translation. Unsupervised pretraining techniques, denoising autoencoders, back translation, and shared latent representation mechanisms are used to simulate the translation task using just monolingual corpora. This paper uses pseudo-parallel data to construct unsupervised neural machine translation (NMT) and dissimilar language pair analysis. This paper firstly analyzes the low performance of unsupervised translation on dissimilar language pairs from three aspects: bilingual word embedding quality, shared words, and word order. And artificial shared word replacement and preordering strategies are proposed to increase the shared words between dissimilar language pairs and reduce the difference in their syntactic structure, thereby improving the translation performance on dissimilar language pairs. The denoising autoencoder and shared latent representation mechanism in unsupervised English machine translation are only required in the early stage of training, and learning the shared latent representation limits the further improvement of performance in different directions. While training the denoising autoencoder by repeatedly altering the training data slows down the convergence of the model, this is especially true for divergent languages. This paper presents an unsupervised NMT model based on pseudo-parallel data to address this issue. It trains two standard supervised neural machine translation models using the pseudo-parallel corpus generated by the unsupervised neural machine translation system, which enhances translation performance and speeds convergence. Finally, the English intelligent translation model is deployed in the wireless network server, and users can access it through the wireless network.

Психофізіологічні характеристики спортсменів як чинник системи підготовки у водному поло

At the present stage of sports development the intensity and volume of training sessions have increased to physiological maximum, requirements for technical and tactical training of highly qualified players and the duration of their stay in the sports arena have increased, so coaches, scientists and doctors are looking for reserves to improve performance. In recent years, differentiated approaches to the training of athletes at different stages of long-term training, taking into account the psychophysiological characteristics of athletes, have been successfully implemented in various sports. Given these trends, there is an urgent question of approach to technical and tactical training of water polo players, taking into account their psychophysiological characteristics. The purpose of the study was to determine the role of individual psychophysiological characteristics in the system of training athletes in water polo. Materials and methods. In the course of the research, we used the methods of theoretical analysis and generalization of literary sources, documentary method. In order to conduct a survey of water polo coaches, we conducted a survey. In order to analyze and process the results of the questionnaire, we used the methods of mathematical statistics. Results and discussion. To conduct a survey of water polo coaches, we developed a questionnaire that included questions about the specifics of water polo training, identifying the most talented water polo players in the early stages of long-term training and psychophysiological characteristics in the water polo training system. In the course of the research, we conducted a survey of water polo coaches on the peculiarities of taking into account the individual psychophysiological characteristics of athletes in the system of training in water polo. As a result of the survey, it was determined that psychophysiological characteristics are an important component of the system of training athletes in water polo. The most significant psychophysiological characteristics for water polo players of different game roles are determined. Conclusion. The study found that all coaches we interviewed (100%) believe it is necessary to take into account the psychophysiological characteristics of athletes in the selection and construction of training athletes in water polo. Psychophysiological characteristics of athletes are one of the most informative criteria for gifted athletes in water polo in the early stages of long-term training, according to 43% of respondents, along with anthropometric data of athletes (14%) and motivation for water polo (29%). As a result of the survey, respondents noted the most effective psychophysiological characteristics for different game roles, in particular for goalkeepers it is the speed of simple and complex motor reactions, for defenders – the ability to anticipate. For midfielders it is speed of thinking and decision making and for attackers it is speed of motor reactions and speed of thinking

Тактична підготовка в командних ігрових видах спорту на ранніх етапах багаторічного удосконалення спортсменів: постановка проблеми

The purpose of the study was to identify the importance and features of tactical training of athletes in team games in the early stages of long-term sports development. Materials and methods. The most common team game sports are singled out, in which the main game tools are the ball, namely: volleyball, handball, basketball and football. Curricula for children's and youth sports school and others are analyzed in each sport in order to identify the place and features of tactical training of athletes in the early stages of long-term sports improvement. Also, Ukrainian and foreign basic works in the field of sports and specialized literature for each of the above sports were developed. Results and discussion. Tactical training of the athlete is one of the determining factors in the effectiveness of his/her performances during competitive activities in team sports (football, volleyball, basketball, handball, etc.). At the same time, in sports theory to this day, it is considered that tactical training is a priority in the preparation of later stages of long-term improvement. In contrast, some domestic and foreign experts in the field of sports argue that athletes who are at the stage of preliminary basic training are already fully capable of effective implementation of tactical tasks during training and competitive activities in the chosen sport. In Ukraine, most children who specialize in team sports are involved in state sports organizations. When building a long-term process of training athletes, these organizations are guided by approved training programs in sports. Most of these programs are outdated and cannot provide effective training for Ukrainian athletes. One of the main conditions for the active development of team sports abroad is a change in approaches to building the educational process of young athletes. In fact, the content of these approaches is to increase the overall share of tactical training within the training process of young athletes, through the active use of games and competitive exercises. This is actively followed in the training of young players in leading European and world private clubs. The use of modern approaches to the training of young athletes in team sports in Ukraine will improve their level of training (including tactical). Conclusion. The study found that tactical training is an integral part of the training process of young athletes in team sports. However, the role of tactical training in the early stages of long-term training is insignificant. The need to update curricula to improve the training of young Ukrainian athletes in the early stages of long-term sports development has been identified. Also, differences between the theoretical and practical support of the process of tactical training of young athletes in Ukraine are identified

Quality assessment of anatomical MRI images from generative adversarial networks: Human assessment and image quality metrics

BackgroundGenerative Adversarial Networks (GANs) can synthesize brain images from image or noise input. So far, the gold standard for assessing the quality of the generated images has been human expert ratings. However, due to limitations of human assessment in terms of cost, scalability, and the limited sensitivity of the human eye to more subtle statistical relationships, a more automated approach towards evaluating GANs is required. New methodWe investigated to what extent visual quality can be assessed using image quality metrics and we used group analysis and spatial independent components analysis to verify that the GAN reproduces multivariate statistical relationships found in real data. Reference human data was obtained by recruiting neuroimaging experts to assess real Magnetic Resonance (MR) images and images generated by a GAN. Image quality was manipulated by exporting images at different stages of GAN training. ResultsExperts were sensitive to changes in image quality as evidenced by ratings and reaction times, and the generated images reproduced group effects (age, gender) and spatial correlations moderately well. We also surveyed a number of image quality metrics. Overall, Fréchet Inception Distance (FID), Maximum Mean Discrepancy (MMD) and Naturalness Image Quality Evaluator (NIQE) showed sensitivity to image quality and good correspondence with the human data, especially for lower-quality images (i.e., images from early stages of GAN training). However, only a Deep Quality Assessment (QA) model trained on human ratings was able to reproduce the subtle differences between higher-quality images. ConclusionsWe recommend a combination of group analyses, spatial correlation analyses, and both distortion metrics (FID, MMD, NIQE) and perceptual models (Deep QA) for a comprehensive evaluation and comparison of brain images produced by GANs.