Training Criteria Research Articles

Problems of professional self-realization of a young scientist in the conditions of transformation of institutes of educational and scientific institutions

Introduction. The article presents the problem of professional self-realization and self-determination of a young scientist in the conditions of transformation of educational and scientific institutions. The sharpness of it is added by the diversification of professional guidelines, supported by the vagueness of state policy in the field of fundamental science development, as well as its orientation to the use of quantitative assessment parameters.Purpose setting. The problem of professional self-determination and self-realization of a young scientist is considered not through the prism of postgraduate training and quantitative criteria for evaluating its effectiveness, but through the analysis and comparison of the dominant motives for carrying out research activities by young scientists, both in Russia and abroad, as well as the factors that determine its formation. This angle of the study represents the novelty of the work.The research methodology is based on the application of structural-functional and institutional approaches, as well as theories that reveal the process of self-determination, self-realization and motivation.Results. The author provides a wide range of examples of studying the phenomenon of self-realization in the scientific environment, both by Russian and foreign researchers. It shows the general and particular that distinguishes the process in Russia and other countries.Conclusions. The main problem of self-determination and self-realization is not so much the satisfaction of material needs as the strengthening of the meaning- forming component of research activity. It should be supported by transparency of criteria focused on the evaluation of the process, the development of the professional environment of interaction, including through improving the social status of scientists representing the middle age group. In this case, it will create prerequisites for young researchers to be confident in the future in science and education.

Characterizing the Accessibility and Content of Surgical and Medical Retina Fellowship Program Websites.

Purpose: To assess the accessibility and content of surgical and medical retina fellowship websites. Methods: The websites of all surgical and medical retina fellowship programs were examined. Each program's website was evaluated based on information available on 10 recruitment and 10 training criteria. The presence of the criteria was summed to calculate a total content score (range, 0-20). Also examined were the differences in website content score by number of fellows, geographic location, and Association of University Professors of Ophthalmology (AUPO) compliance. Results: This study identified 102 surgical and 25 medical retina programs. Overall, 91.2% of surgical and 88.0% of medical retina programs had an accessible website. The surgical retina program website contained a mean of 9.8 of the total criteria, including 4.9 recruitment criteria and 5.2 training criteria, with no significant differences by number of fellows, geography, or AUPO status. Medical retina websites contained a mean of 9.3 total criteria, including 4.5 recruitment criteria and 4.9 training criteria. Website content scores for medical retina programs were associated with geography and AUPO status, which was consistent when stratifyed by recruitment and training criteria. Conclusions: Most surgical and medical retina fellowships have an accessible program website. However, there are opportunities to improve the completeness and consistency of information on these websites. Improved websites can help programs attract well-suited candidates and might address multiple inefficiencies in the application process.

Essential integrated behavioral health care skills for primary care providers: A Delphi study.

Workforce development is a vital process for advancing the field of integrated behavioral health in primary care (IBHPC) and depends on the development of job-specific skills that meet the needs of patients and other team members. Job-specific skills guide both university- and worksite-based training programs in preparing future cohorts of health care professionals in IBHPC; however, there are no extant training criteria for nontrainee primary care providers. A multiround survey using a modified Delphi technique was conducted. Participants were recruited from a large national research network, university alumni and community health center partners. Participants included allopathic and osteopathic physicians, physician assistants and nurse practitioners serving as primary care providers in a variety of IBHPC settings. In the survey, participants were instructed to categorize 22 skills as "Essential," "Compatible," or "Irrelevant" for IBHPC practice. Categorized responses were analyzed using descriptive statistics and comments were analyzed using a modified thematic analysis method. Fifty-four participants were recruited for the study, with 43 participating in round 1, 38 participating in round 2, and 36 participating in round 3. After 3 rounds, 19 skills received at least 80% consensus as "Essential" and 22 received at least 50% consensus as "Essential." Comment themes revealed participants were overall agreeable with skills, but had concerns about role, training, and time constraints. Primary care providers can prepare for and enhance IBHPC practice by developing skills identified as essential. Employers can adapt this skill list for efforts related to recruiting, hiring, and onboarding. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Basil: A Fast and Byzantine-Resilient Approach for Decentralized Training

Detection and mitigation of Byzantine behaviors in a decentralized learning setting is a daunting task, especially when the data distribution at the users is heterogeneous. As our main contribution, we propose Basil, a fast and computationally efficient Byzantine robust algorithm for decentralized training systems, which leverages a novel sequential, memory assisted and performance-based criteria for training over a logical ring while filtering the Byzantine users. In the IID dataset distribution setting, we provide the theoretical convergence guarantees of Basil, demonstrating its linear convergence rate. Furthermore, for the IID setting, we experimentally demonstrate that Basil is robust to various Byzantine attacks, including the strong Hidden attack, while providing up to ${\sim}16 \%$ higher test accuracy over the state-of-the-art Byzantine-resilient decentralized learning approach. Additionally, we generalize Basil to the non-IID dataset distribution setting by proposing Anonymous Cyclic Data Sharing (ACDS), a technique that allows each node to anonymously share a random fraction of its local non-sensitive dataset (e.g., landmarks images) with all other nodes. We demonstrate that Basil alongside ACDS with only $5\%$ data sharing provides effective toleration of Byzantine nodes, unlike the state-of-the-art Byzantine robust algorithm that completely fails in the heterogeneous data setting. Finally, to reduce the overall latency of Basil resulting from its sequential implementation over the logical ring, we propose Basil+. In particular, Basil+ provides scalability by enabling Byzantine-robust parallel training across groups of logical rings, and at the same time, it retains the performance gains of Basil due to sequential training within each group. Furthermore, we experimentally demonstrate the scalability gains of Basil+ through different sets of experiments.

Equines do not live for grass alone: Training equines using social interaction as a reinforcer

Functional analysis data and previous studies on animal training have demonstrated that social interaction with humans can serve as a reinforcer for animals. Yet, some studies have demonstrated that tactile interaction (e.g., patting, petting, or scratching) is less effective or ineffective when compared to food. However, the reinforcement procedures used may account for these discrepancies. The current study investigated whether tactile interaction, in the form of petting and scratching, could be used as a reinforcer to train behaviors to two horses and a mule. First, each equine learned when reinforcement would be available and what behaviors to engage in during reinforcement delivery. Next, a series of shaping steps and a changing-criterion design were used to test whether tactile interaction could be used to shape two new behaviors, stay and come. All three equines completed reinforcement training and met the mastery criteria for training stay and come. These results demonstrate that tactile interaction can be used as a reinforcer to train equines and also suggest that details of the reinforcement delivery process may be an important consideration when tactile interaction is used as a reinforcer.

Nothing but a “g” Thing? Toward an Integrative Model of Frame-of-Reference Training Effectiveness

Rater training is important for improving the quality of human resource decisions. However, there is little research that has examined the influence of trainee individual differences on rater training effectiveness. Borrowing from self-regulation theory and executive function literature, we developed an individual differences model of rater training effectiveness across three studies. In Study 1, we found support for a partially mediated model in which learning goal orientation, internal locus of control, and promotion focus influenced rater training motivation through general self-efficacy. In Study 2, we extended the model to include cognitive ability. Study 3 further extended the model to include rater training criteria, and we found that rater training motivation did not predict declarative knowledge, rating validity, or rating accuracy, but cognitive ability predicted all three outcomes. Overall, the results suggest that although individual differences in attitudes are relevant to understanding rater training motivation, cognitive ability may be much more important for understanding rater training effectiveness. Implications for rater training theory and practice are discussed.

Application of machine learning models based on decision trees in classifying the factors affecting mortality of COVID-19 patients in Hamadan, Iran

BackgroundDue to the high mortality of COVID-19 patients, the use of a high-precision classification model of patient’s mortality that is also interpretable, could help reduce mortality and take appropriate action urgently. In this study, the random forest method was used to select the effective features in COVID-19 mortality and the classification was performed using logistic model tree (LMT), classification and regression tree (CART), C4.5, and C5.0 tree based on important features.MethodsIn this retrospective study, the data of 2470 COVID-19 patients admitted to hospitals in Hamadan, west Iran, were used, of which 75.02% recovered and 24.98% died. To classify, at first among the 25 demographic, clinical, and laboratory findings, features with a relative importance more than 6% were selected by random forest. Then LMT, C4.5, C5.0, and CART trees were developed and the accuracy of classification performance was evaluated with recall, accuracy, and F1-score criteria for training, test, and total datasets. At last, the best tree was developed and the receiver operating characteristic curve and area under the curve (AUC) value were reported.ResultsThe results of this study showed that among demographic and clinical features gender and age, and among laboratory findings blood urea nitrogen, partial thromboplastin time, serum glutamic-oxaloacetic transaminase, and erythrocyte sedimentation rate had more than 6% relative importance. Developing the trees using the above features revealed that the CART with the values of F1-score, Accuracy, and Recall, 0.8681, 0.7824, and 0.955, respectively, for the test dataset and 0.8667, 0.7834, and 0.9385, respectively, for the total dataset had the best performance. The AUC value obtained for the CART was 79.5%.ConclusionsFinding a highly accurate and qualified model for interpreting the classification of a response that is considered clinically consequential is critical at all stages, including treatment and immediate decision making. In this study, the CART with its high accuracy for diagnosing and classifying mortality of COVID-19 patients as well as prioritizing important demographic, clinical, and laboratory findings in an interpretable format, risk factors for prognosis of COVID-19 patients mortality identify and enable immediate and appropriate decisions for health professionals and physicians.

Multi-Stakeholder Research Data Management Training as a Tool to Improve the Quality, Integrity, Reliability and Reproducibility of Research

To ensure the quality and integrity of data and the reliability of research, data must be well documented, organised, and described. This calls for research data management (RDM) education for researchers. In light of 3 ECTS Basics of Research Data Management (BRDM) courses held between 2019 and 2021, we aim to find how a generic level multi-stakeholder training can improve STEM and HSS disciplines’ doctoral students’ and postdoc researchers’ competencies in RDM. The study uses quantitative, descriptive and inferential statistics to analyse respondents’ self-ratings of their competencies, and a qualitative grounded theory-inspired approach to code and analyse course participants’ feedback. Results: On average, based on the post-course surveys, respondents’ (n = 123) competencies improved one point on a four-level scale, from “little competence” (2) to “somewhat competent” (3). Participants also reported that the training would change their current practices in planning research projects, data management and documentation, acknowledging legal and data privacy viewpoints, and data collecting and organising. Participants indicated that it would be helpful to see legal and data privacy principles and regulations presented as concrete instructions, cases, and examples. The most requested continuing education topics were metadata and description, discipline specific cultures, and backup, version management, and storage. Conclusions: Regarding to the widely used criteria for successful training containing 1) active participation during training; 2) demand for RDM training; 3) increased participants’ knowledge and understanding of RDM and confidence in enacting RDM practices; and 4) positive post-training feedback, BRDM meets the criteria. This study shows that although reaching excellent competence in a RDM basics training is improbable, participants become aware of RDM and its contents and gain the elementary tools and basic skills to begin applying sound RDM practices in their research. Furthermore, participants are introduced to the academic and research support professionals and vice versa: Stakeholders will get to know the challenges that young researchers and research students encounter when applying RDM. The study reveals valuable information on doctoral students’ and postdoc researchers’ competencies, the impact of education on competencies, and further learning needs in RDM.

Determination of Model Characteristics and Evaluation Criteria for Strength Training of Qualified Armwrestling Student-Athletes

Research purpose. The aim of the study was to determine the model indicators of strength readiness of qualified armwrestling student-athletes in the weight category from 80 to 100 kg in 4 strength exercises: flexion of the fingers, stretch with a hammer, hook and bending of the hand.
 Materials and Methods. The study involved 16 qualified armwrestling student-athletes weighing from 80 to 100 kg (M = 86,63 kg), 7 of which are participants of national championships of Ukraine, and 9 of which are prize-winners of competitions of regional level. During the study, strength indicators were determined in 4 competitive exercises. Strength indices in all test exercises were measured in a static mode by an FB5k series digital force gauge (Poland) with an accuracy class of up to 100 g, which was mounted on a special armwrestling table using a specially made block device. The created design was called the “ARM1 Device” (patent 43082). During the statistical analysis, the following parameters were calculated: arithmetic mean (M), standard deviation (σ), error of representativeness (m) and the coefficient of variation (V).
 Results. The analysis of the strength capabilities of the qualified student-armwrestlers and the presented models of their strength readiness made it possible to develop assessment criteria for the level of strength readiness of the qualified student-armwrestlers. Three levels of strength readiness have been identified: low (M - σ), high (M + σ) and medium, i.e. more than (M - σ) but less than (M + σ). Comparative analysis of the ratios of different levels of strength training in this group of armwrestling athletes allowed to determine from 128 indicators of strength 14.84% (n = 19) of low level, 62.50% (n = 80) of medium level and 22.66% (n = 29) of high level.
 Conclusions. The study made it possible to establish the features of the levels of development of strength training of qualified armwrestling athletes weighing from 80 to 100 kg. The model indicators of strength training of qualified armwrestlers were also determined. The analysis found that the subject athletes have an insufficient level of strength training and are not fully realized in terms of strength capabilities (62.50% indicators of the average level of strength).

Multisample Online Learning for Probabilistic Spiking Neural Networks.

Spiking neural networks (SNNs) capture some of the efficiency of biological brains for inference and learning via the dynamic, online, and event-driven processing of binary time series. Most existing learning algorithms for SNNs are based on deterministic neuronal models, such as leaky integrate-and-fire, and rely on heuristic approximations of backpropagation through time that enforces constraints such as locality. In contrast, probabilistic SNN models can be trained directly via principled online, local, and update rules that have proven to be particularly effective for resource-constrained systems. This article investigates another advantage of probabilistic SNNs, namely, their capacity to generate independent outputs when queried over the same input. It is shown that the multiple generated output samples can be used during inference to robustify decisions and to quantify uncertainty-a feature that deterministic SNN models cannot provide. Furthermore, they can be leveraged for training in order to obtain more accurate statistical estimates of the log-loss training criterion and its gradient. Specifically, this article introduces an online learning rule based on generalized expectation-maximization (GEM) that follows a three-factor form with global learning signals and is referred to as GEM-SNN. Experimental results on structured output memorization and classification on a standard neuromorphic dataset demonstrate significant improvements in terms of log-likelihood, accuracy, and calibration when increasing the number of samples used for inference and training.

Training Needs Assessment on Blended Learning Model Management to Improve Teachers’ Competence

This research is motivated by the existence of gaps and problems in the low competence of teachers in Indonesia. The implementation of blended learning model training is a solution to improve teacher competence. The training will run effectively, efficiently as the target if it can implement training management properly. The first stage of training management is Training Needs Assessment. Training Needs Assessment (TNA) is the first stage in training management that determines whether the next stage is successful or not. TNA is also a diagnostic process in order to be able to determine/take pictures of the problems that are currently being faced and how the challenges will be in the future and both of these things that must be fulfilled in the implementation of the training. The purpose of this study is to describe and analyze TNA in blended learning model training management in improving teacher competence at the Center for the Development and Empowerment of Science Educators and Educators (PPPPTK) Bandung, West Java. The research approach uses a qualitative approach with a case study method. The results showed that PPPPTK IPA in implementing blended learning model training management in improving teacher competence had implemented TNA with the stages of Analyze training needs and Identify training objectives and criteria.

Training Needs Assessment on Blended Learning Model Management to Improve Teachers’ Competence

This research is motivated by the existence of gaps and problems in the low competence of teachers in Indonesia. The implementation of blended learning model training is a solution to improve teacher competence. The training will run effectively, efficiently as the target if it can implement training management properly. The first stage of training management is Training Needs Assessment. Training Needs Assessment (TNA) is the first stage in training management that determines whether the next stage is successful or not. TNA is also a diagnostic process in order to be able to determine/take pictures of the problems that are currently being faced and how the challenges will be in the future and both of these things that must be fulfilled in the implementation of the training. The purpose of this study is to describe and analyze TNA in blended learning model training management in improving teacher competence at the Center for the Development and Empowerment of Science Educators and Educators (PPPPTK) Bandung, West Java. The research approach uses a qualitative approach with a case study method. The results showed that PPPPTK IPA in implementing blended learning model training management in improving teacher competence had implemented TNA with the stages of Analyze training needs and Identify training objectives and criteria.

HISTOPATHOLOGY IMAGE CLASSIFICATION USING HYBRID PARALLEL STRUCTURED DEEP-CNN MODELS

The healthcare industry is one of the many out there that could majorly benefit from advancement in the technology it utilizes. Artificial intelligence (AI) technologies are especially integral and specifically deep learning (DL); a highly useful data-driven technology. It is applied in a variety of different methods but it mainly depends on the structure of the available data. However, with varying applications, this technology produces data in different contexts with particular connotations. Reports which are the images of scans play a great role in identifying the existence of the disease in a patient. Further, the automation in processing these images using technology like CNN-based models makes it highly efficient in reducing human errors otherwise resulting in large data. Hence this study presents a hybrid deep learning architecture to classify the histopathology images to identify the presence of cancer in a patient. Further, the proposed models are parallelized using the TensorFlow-GPU framework to accelerate the training of these deep CNN (Convolution Neural Networks) architectures. This study uses the transfer learning technique during training and early stopping criteria are used to avoid overfitting during the training phase. these models use LSTM parallel layer imposed in the model to experiment with four considered architectures such as MobileNet, VGG16, and ResNet with 101 and 152 layers. The experimental results produced by these hybrid models show that the capability of Hybrid ResNet101 and Hybrid ResNet152 architectures are highly suitable with an accuracy of 90% and 92%. Finally, this study concludes that the proposed Hybrid ResNet-152 architecture is highly efficient in classifying the histopathology images. The proposed study has conducted a well-focused and detailed experimental study which will further help researchers to understand the deep CNN architectures to be applied in application development.

The need for individualizing teaching and assurance of knowledge transmission to patients training for home dialysis.

Patients have varied learning styles and this has implications for home haemodialysis (HHD). Assessment tools directed toward understanding these styles remains understudied. As a consequence, this may lead to substandard retention rates or adverse events in HHD programs. As part of a continuous quality improvement initiative we have aimed to improve our understanding of patient learning styles and consequently tailor home dialysis training to individuals. To objectively determine knowledge translation and comprehension, irrespective of learning styles, we have introduced an objective structured clinical examination (OSCE). This assessment tool allows for further refinement of educational priorities by highlighting both deficiencies and strengths. Thereafter, an exit OSCE ensures patients attain an acceptable standard to complete home haemodialysis independently. We hope this tool will help shape future training criteria for HHD programs and consequently reduce adverse event rates.

Epistemic stability and epistemic adaptability: interdisciplinary knowledge integration competencies for complex sustainability issues

Global sustainability challenges transcend disciplines and, therefore, demand interdisciplinary approaches that are characterized by cross-disciplinary collaboration and integration across disciplines. In accordance with this need for interdisciplinary approaches, sustainability professionals have been reported to require interdisciplinary competencies. Although the necessity of interdisciplinary competencies is generally agreed upon, and there has been extensive research to understand competencies for interdisciplinarity, there is still no comprehensive understanding of how individual competencies shape the ability to integrate knowledge across disciplines. Therefore, based on empirical research and literature review, we propose a novel framework to understand competencies for interdisciplinarity. The empirical data were collected through written reflection and interviews with 19 students in the context of an interdisciplinary master’s course. We describe four typical behaviours—naïve, assertive, accommodating, and integrative. Based on these behavioural typologies, we define two sets of competencies that collaborators require to engage in interdisciplinary knowledge integration: Epistemic Stability (ES) and Epistemic Adaptability (EA). ES competencies are the competencies to contribute one’s own academic knowledge, such as theoretical and methodological grounding in one’s own field and confidence, and EA competencies are the competencies to engage with academic knowledge contributed by others, such as curiosity, openness and communicative skills. Our findings show that interdisciplinary knowledge integration requires ES and EA competencies. Our framework for interdisciplinary competencies offers insights for revising and designing more interventions to prepare (future) professionals for interdisciplinary work on sustainability issues, providing insights on criteria for assessment, management, and training.

Metrological approach to the assessment of the criteria of speed and strength training of middle school students in physical education lessons in athletics

Evaluation of the results of pedagogical control is the main factor for obtaining information about the state of physical fitness and effectiveness of educational work in physical education lessons. Obtaining an objective assessment of the level of development of speed and strength qualities of middle school students helps to improve the management of the educational process, making adjustments in the planning of physical education and forecasting academic achievement. The main metrological requirements for the development of evaluation criteria for the manifestation of speed and strength training of students include: 1) the correct choice of control, tests and their compliance with statistical criteria of reliability, objectivity and informativeness; 2) compliance of methods and means of control with the objectives of the study; 3) the adequacy of methods of statistical processing of test results. The purpose of the study is aimed at developing assessment criteria for speed and strength training of middle school students in physical education classes in athletics. The research methodology is aimed at substantiating the results of pedagogical control, development of quantitative and qualitative criteria of speed and strength training of middle school students of general secondary schools. The scientific novelty is to develop quantitative and qualitative criteria for assessing the speed and strength of middle school students, to determine the impact of the process of physical education on the dynamics of their changes in the results of phased control. The use of metrological tools and methods for assessing the speed and strength of middle school students allows the implementation of individual and differential approach to each student in the process of physical education.

Prospect and challenges of analog switching for neuromorphic hardware

To inaugurate energy-efficient hardware as a solution to complex tasks, information processing paradigms shift from von Neumann to non-von Neumann computing architectures. Emerging electronic devices compete with speed, energy, and performance to revolutionize the neural hardware system where training and inference must achieve milestones. In this Perspective, we discuss the essential criteria for training and inference in various nonvolatile neuromorphic systems such as filamentary resistive switching, interfacial resistive switching, electrochemical random-access memory, and ferroelectric memory. We present a holistic analysis of technical requirements to design ideal neuromorphic hardware in which linearity is the critical aspect during training, whereas retention is the essential criterion of inference. Finally, we evaluate the prospect of a futuristic neuromorphic hardware system by optimizing the training and inference dilemma.

Multi-Channel Talker-Independent Speaker Separation Through Location-Based Training

Permutation ambiguity is a crucial issue for deep learning based talker-independent speaker separation. Deep clustering and permutation invariant training (PIT) have been widely used to address the permutation ambiguity problem in monaural scenarios. Although both approaches have been extended to multi-microphone scenarios, we believe that the permutation ambiguity problem can be naturally avoided by leveraging the spatial relations of multiple speakers. In this study, we present location-based training (LBT), a new approach to achieve talker independency in multi-channel speaker separation. Unlike PIT that examines all possible permutations, LBT assigns speakers according to their positions in physical space. With a linear training complexity to the number of concurrent speakers, LBT is computationally much more efficient than PIT with a factorial complexity, particularly when a large number of overlapping speakers needs to be separated. Specifically, we propose two training criteria: azimuth-based and distance-based training, using speaker azimuths and distances relative to a microphone array, respectively. Evaluation results show that LBT significantly outperforms PIT on two-speaker and three-speaker mixtures with different array geometries and in various acoustic conditions. In addition, we propose a joint training strategy to integrate azimuth-based and distance-based training, which further improves separation performance.

Using Pareto simulated annealing to address algorithmic bias in machine learning

Abstract Algorithmic bias arises in machine learning when models that may have reasonable overall accuracy are biased in favor of ‘good’ outcomes for one side of a sensitive category, for example gender or race. The bias will manifest as an underestimation of good outcomes for the under-represented minority. In a sense, we should not be surprised that a model might be biased when it has not been ‘asked’ not to be; reasonable accuracy can be achieved by ignoring the under-represented minority. A common strategy to address this issue is to include fairness as a component in the learning objective. In this paper, we consider including fairness as an additional criterion in model training and propose a multi-objective optimization strategy using Pareto Simulated Annealing that optimizes for both accuracy and underestimation bias. Our experiments show that this strategy can identify families of models with members representing different accuracy/fairness tradeoffs. We demonstrate the effectiveness of this strategy on two synthetic and two real-world datasets.

A Comprehensive Study on Deep Learning-Based Methods for Sign Language Recognition

In this paper, a comparative experimental assessment of computer vision-based methods for sign language recognition is conducted. By implementing the most recent deep neural network methods in this field, a thorough evaluation on multiple publicly available datasets is performed. The aim of the present study is to provide insights on sign language recognition, focusing on mapping non-segmented video streams to glosses. For this task, two new sequence training criteria, known from the fields of speech and scene text recognition, are introduced. Furthermore, a plethora of pretraining schemes is thoroughly discussed. Finally, a new RGB+D dataset for the Greek sign language is created. To the best of our knowledge, this is the first sign language dataset where sentence and gloss level annotations are provided for a video capture.