Categorization Training Research Articles

Zero-Shot Cross-Media Embedding Learning With Dual Adversarial Distribution Network

Existing cross-media retrieval methods are mainly based on the condition where the training set covers all the categories in the testing set, which lack extensibility to retrieve data of new categories. Thus, zero-shot cross-media retrieval has been a promising direction in practical application, aiming to retrieve data of new categories (unseen categories), only with data of limited known categories (seen categories) for training. It is challenging for not only the heterogeneous distributions across different media types, but also the inconsistent semantics across seen and unseen categories need to be handled. To address the above issues, we propose dual adversarial distribution network (DADN) , to learn common embeddings and explore the knowledge from word-embeddings of different categories. The main contributions are as follows. First, zero-shot cross-media dual generative adversarial networks architecture is proposed, in which two kinds of generative adversarial networks (GANs) for common embedding generation and representation reconstruction form dual processes. The dual GANs mutually promote to model semantic and underlying structure information, which generalizes across different categories on heterogeneous distributions and boosts correlation learning. Second, distribution matching with maximum mean discrepancy criterion is proposed to combine with dual GANs, which enhances distribution matching between common embeddings and category word-embeddings. Finally, adversarial inter-media metric constraint is proposed with an inter-media loss and a quadruplet loss, which further model the inter-media correlation information and improve semantic ranking ability. The experiments on four widely used cross-media datasets demonstrate the effectiveness of our DADN approach.

Predicting Fire Hazard Areas Using Vegetation Indexes, Case Study: Forests of Golestan Province, Iran

Every year, hundreds of fires occur in the forests and rangelands across the world and damage thousands hectare of trees, shrubs, and plants which cause environmental and economical damages. This study aims to establish a real time forest fire alert system for better forest management and monitoring in Golestan Province. In this study, in order to prepare fire hazard maps, the required layers were produced based on fire data in Golestan forests and MODIS sensor data.At first, the natural fire data was divided into two categories of training and test samples randomly. Then, the vegetation moisture stresses and greenness were considered using six indexes of NDVI, MSI, WDVI, OSAVI, GVMI and NDWI in natural fire area of training category on the day before fire occurrence and a long period of 15 years, and the risk threshold of the parameters was considered in addition to selecting the best spectral index of vegetation. Finally, the model output was validated for fire occurrences of the test category. The results showed the possibility of prediction of fire site before occurrence of fire with more than 80 percent accuracy.

Novel representations that support rule-based categorization are acquired on-the-fly during category learning.

Humans learn categorization rules that are aligned with separable dimensions through a rule-based learning system, which makes learning faster and easier to generalize than categorization rules that require integration of information from different dimensions. Recent research suggests that learning to categorize objects along a completely novel dimension changes its perceptual representation, making it more separable and discriminable. Here, we asked whether such newly learned dimensions could support rule-based category learning. One group received extensive categorization training and a second group did not receive such training. Later, both groups were trained in a task that made use of the category-relevant dimension, and then tested in an analogical transfer task (Experiment 1) and a button-switch interference task (Experiment 2). We expected that only the group with extensive pre-training (with well-learned dimensional representations) would show evidence of rule-based behavior in these tasks. Surprisingly, both groups performed as expected from rule-based learning. A third experiment tested whether a single session (less than 1 h) of training in a categorization task would facilitate learning in a task requiring executive function. There was a substantial learning advantage for a group with brief pre-training with the relevant dimension. We hypothesize that extensive experience with separable dimensions is not required for rule-based category learning; rather, the rule-based system may learn representations "on the fly" that allow rule application. We discuss what kind of neurocomputational model might explain these data best.

Mental health training, attitudes toward support, and screening positive for mental disorders

ABSTRACTPublic Safety Personnel (PSP; e.g. correctional workers, dispatchers, firefighters, paramedics, police) are frequently exposed to potentially traumatic events (PTEs). Several mental health training program categories (e.g. critical incident stress management (CISM), debriefing, peer support, psychoeducation, mental health first aid, Road to Mental Readiness [R2MR]) exist as efforts to minimize the impact of exposures, often using cognitive behavioral therapy model content, but with limited effectiveness research. The current study assessed PSP perceptions of access to professional (i.e. physicians, psychologists, psychiatrists, employee assistance programs, chaplains) and non-professional (i.e. spouse, friends, colleagues, leadership) support, and associations between training and mental health. Participants included 4,020 currently serving PSP participants. Data were analyzed using cross-tabulations and logistic regressions. Most PSP reported access to professional and non-professional support; nevertheless, most would first access a spouse (74%) and many would never, or only as a last resort, access professional support (43–60%) or PSP leaders (67%). Participation in any mental health training category was associated with lower (p < .01) rates for some, but not all, mental disorders, with no robust differences across categories. Revisions to training programs may improve willingness to access professional support; in the interim, training and support for PSP spouses and leaders may also be beneficial.

BioReader: a text mining tool for performing classification of biomedical literature

BackgroundScientific data and research results are being published at an unprecedented rate. Many database curators and researchers utilize data and information from the primary literature to populate databases, form hypotheses, or as the basis for analyses or validation of results. These efforts largely rely on manual literature surveys for collection of these data, and while querying the vast amounts of literature using keywords is enabled by repositories such as PubMed, filtering relevant articles from such query results can be a non-trivial and highly time consuming task.ResultsWe here present a tool that enables users to perform classification of scientific literature by text mining-based classification of article abstracts. BioReader (Biomedical Research Article Distiller) is trained by uploading article corpora for two training categories - e.g. one positive and one negative for content of interest - as well as one corpus of abstracts to be classified and/or a search string to query PubMed for articles. The corpora are submitted as lists of PubMed IDs and the abstracts are automatically downloaded from PubMed, preprocessed, and the unclassified corpus is classified using the best performing classification algorithm out of ten implemented algorithms.ConclusionBioReader supports data and information collection by implementing text mining-based classification of primary biomedical literature in a web interface, thus enabling curators and researchers to take advantage of the vast amounts of data and information in the published literature. BioReader outperforms existing tools with similar functionalities and expands the features used for mining literature in database curation efforts. The tool is freely available as a web service at http://www.cbs.dtu.dk/services/BioReader

Analysis and classification of the road traffic health and safety mobile apps based on the Haddon’s matrix

Background: Road traffic accidents have been one of the main causes of death worldwide. However, most of the crashes are both predictable and preventable. With the widespread use, mobile phone is considered as a major road safety risk. However, varieties of apps have recently been developed for improving the road traffic health and safety. Despite an increasing trend of these apps, there is no comprehensive analysis of their features and no taxonomy or classification of them based on the traffic safety theories or frameworks. The aim of this study was to explore characteristics of available mobile apps on the road traffic health and safety and classify them based on the Haddon’s Matrix. Methods: To conduct this research, a comprehensive and systematic review of the mobile applications developed for the road traffic health and safety was carried out through using the qualitative content analysis. Google Play was searched through using a combination of the keywords to retrieve the road traffic apps. In order to extract the app features, their description was examined, and their content was analyzed. Then they were classified in four main categories including Road Traffic Health & safety, Road Traffic Training, Road traffic Navigation, and Other Road Traffic apps. Finally, the Haddon’s matrix was applied to analyze and classify those mobile apps residing in two main categories of the traffic health & safety, and the traffic training. Haddon’s matrix is a relevant framework for structured analyses of traffic injury events. In the Haddon’s matrix, the contributions of human, vehicle/equipment and environmental factors to the injuries in the three phases pre-crash, crash and post-crash has been presented. Results: In this study, 916 mobile apps met the inclusion criteria and were included in the final analysis.19 subcategories were identified for classifying the included apps based on their features and functionalities. In total, 620 of the mobile apps were grouped on the basis of Haddon’s matrix. About 61.76% of these apps were categorized as the traffic health & safety group. Among the subcategories of the traffic health & safety group, the highest number of apps (194) was the real-time traffic alerting apps. Behavior based feed-backing subcategory was ranked the second among the traffic health and safety apps with 75 apps. When the researchers applied factors and phases of Haddon’s framework for analyzing the apps, it was found that the Haddon’s factors have been taken into account in the apps grouped into two main categories of the apps including traffic training and traffic health & safety applications. The highest percentage of Haddon’s matrix factor considered for the intervention through the apps was related to the physical & social environment (23.69%) in the traffic training category. The features of 259 apps were classified in the Event/ Driving phase of Haddon’s matrix. Most of the apps in this phase was related to the real-time traffic alerting subgroup. These apps performed in real time while driving. The lowest number of the apps in driving phase was observed in the behavior based feed backing subcategory, which inform drivers about their driving behavior in real-time. Sixteen apps had features that categorized them in both Pre-event / Pre-driving and Event / Driving phases. In addition, 20 apps were found to have feature for intervening in both Event / Driving and Post-event / Post –driving phases. All 235 apps in the traffic training category were in pre-event/ pre-driving phase, which requires users to use these apps before driving. Conclusion: Applying Haddon’s matrix on analyzing and classifying the traffic health and safety apps revealed strengths and weaknesses of the existing related mobile apps in terms of the factors that must be considered for intervention toward prevention and reduction of road traffic accidents as the main public health issue.

Common Ground: Frameworks for Teaching Improvisational Ability in Medical Education

Issue: The practice of medicine is intrinsically unpredictable. Clinicians must respond skillfully to this uncertainty; therefore, medical educators are using improvisational theater training methods to teach improvisational ability in areas such as communication and professionalism. This teaching approach is called “medical improv.” Although early reports of medical improv suggest promise, the collective descriptions of curricular content lack consistency. This ambiguity creates impediments for further implementation and research of this new educational technique. To address this challenge, the author presents two unifying conceptual frameworks for medical improv curricula. Evidence: Medical improv evolved from previous work done in improvisational (improv) theater and “applied improv,” which is the broader category of improv-based training in nontheater environments. The author synthesized curricular elements from all three fields of improv into two conceptual frameworks: core curricular components (skills, principles, and exercises) and core skill groups (attunement, affirmation, and advancement). The curricular component framework simplifies terminology and clarifies a previously vague distinction between skills and principles. The skill group framework harnesses a redundant and wide-ranging list of improv skills into three categories. Together, the frameworks provide a simple structure that encompasses and organizes the core concepts of medical improv. Implications: These curricular frameworks enable educators and learners to more easily understand the scope and depth of medical improv concepts and to identify areas where medical improv could enhance existing medical education strategies. For example, there is substantial overlap between improvisation skills and current curricular objectives for communication skills, as outlined by Accreditation Council for Graduate Medical Education benchmarks (e.g., Milestones, Competencies, Entrustable Professional Activities). Medical improv courses could be an efficient way to achieve many of these objectives. In addition, improv-based exercises represent an experiential learning modality that is underutilized in communication skills training; therefore, exercises could be integrated into existing coursework to bridge the learning pathway between texts and simulation. Furthermore, the deliberate cultivation of affirmation skills could enhance resilience and wellness in clinical and learning environments. In summary, through their conceptual simplicity, these curricular frameworks for medical improv can help educators understand, study, and optimize teaching of improvisation in medical education.

A PECUÁRIA NA FORMAÇÃO SOCIOESPACIAL DE URUGUAIANA/RS / THE LIVESTOCK IN THE SOCIOESPACIAL TRAINING OF URUGUAIANA/RS

A proposta desse trabalho é de compreender como se deu a formação socioespacial do município gaúcho de Uruguaiana através da pecuária. A atividade, que foi inserida em seu território com a chegada dos jesuítas, atravessou diversos momentos e tornou-se a principal atividade econômica do mesmo. Para o entendimento acerca desse processo será utilizado como referencial teórico a categoria de Formação Socioespacial. Essa categoria encaminha-se no sentido de precisar como um dado modo de produção realiza-se nas diversas formações sociais, colaborando no entendimento da atual pecuária extensiva e latifundiária em Uruguaiana.

Gene expression programming models for liquefaction-induced lateral spreading

Lateral spreading is one of the most significant destructive and catastrophic phenomena associated with liquefaction caused by earthquake and can impose very serious damages to structures and engineering facilities. The aim of this study is to evaluate liquefaction induced lateral spreading and finding new relations using gene expression programming (GEP) that is a new and developed generation of genetic algorithms approaches. Since there are complicated, nonlinear and higher order relationships between many factors affecting the lateral spreading, GEP is assumed to be capable of finding complex and accurate relationships between these factors. This study includes three main stages: (i) compilation of available database (484 data), (ii) dividing data into training and testing categories, and (iii) building new models and propose new relationships to predict ground displacement in free face, gentle slope and general ground conditions. The results of modeling each of the different ground conditions are presented in the form of mathematical equations. At the end, the final GEP models for 3 different cases of ground conditions are compared with multiple linear regression (MLR) and other published models. The statistical parameters indicate the higher accuracy of the GEP models over other relations.

Review of Integrated Healthcare Training Series.

Reviews the Integrated Healthcare Training Series, produced by the National Register of Health Service Psychologists (2017). The Integrated Health Care Training Series offers a convenient set of web-based resources for stakeholders of integrated behavioral health (IBH) and is available through the National Register of Health Service Psychologists (NRHSP) website. The series contains 42 videos and over 11 hr of professionally produced lectures classified into six categories: Models and Concepts, Implementation, Practice, Medical Management, Pharmacological Management, and Psychosocial Management. While the first two training categories benefit most IBH stakeholders, the latter four focus on clinical matters and are more appropriate for clinicians. The primary audience for this material is obviously members of the NRHSP (i.e., psychologists); accessibility and appeal to other viewers may be lower. (PsycINFO Database Record (c) 2018 APA, all rights reserved).

Classification of Objects with Occlusions by Means of a DAG-CNN

In the following paper, the training of a Convolutional Neural Network (CNN) type DAG (Directed Acyclic Graph) that classifies 5 tools: scalpel, pliers, scissor, screwdriver and spanner, with a 99.8% accuracy is presented, in order to perform a classification process under unexpected conditions, such as the presence of occlusions, where the visualization of activations of the convolution layers of the network allows to analyze the reasons why the network correctly or incorrectly classifies each image, for cases with different percentages of occlusion on each one of the trained categories. Subsequently, a solution for the classification of occluded objects is proposed, and a new DAG-CNN is trained, with a 99% accuracy, in order to evaluate the feasibility of the said proposal, and to observe the change of the activations against occluding percentages between 0% and 80%. Finally, the results of both networks are compared and the proposal viability and its right conditions are determined.

The impact of training methodology and category structure on the formation of new categories from existing knowledge.

Categorization decisions are made thousands of times every day, and a typical adult knows tens of thousands of categories. It is thus relatively rare that adults learn new categories without somehow reorganizing pre-existing knowledge. Yet, most perceptual categorization research has investigated the ability to learn new categories without considering they relation to existing knowledge. In this article, we test the ability of young adults to merge already known categories into new categories as a function of training methodology and category structures using two experiments. Experiment 1 tests participants' ability to merge rule-based or information-integration categories that are either contiguous, semi-contiguous, or non-contiguous in perceptual space using a classification paradigm. Experiment 2 is similar Experiment 1 but uses a YES/NO learning paradigm instead. The results of both experiments suggest a strong effect of the contiguity of the merged categories in perceptual space that depends on the type of category representation that is learned. The type of category representation that is learned, in turn, depends on a complex interaction of the category structures and training task. We conclude by discussing the relevance of these results for categorization outside the laboratory.

Comparative Study of Medical Reference and Information Mobile Apps for Healthcare Professionals and Students.

This paper presents a framework upon which medical applications can be evaluated both in terms of basic functionality, and their target area of users. The study was conducted on 40 mobile applications that were published within Google Play and Apple App store targeted for clinicians, medical practitioners and students. It was important to first classify the mobile apps selected because there are many generic applications, and some focus on special areas within the medical field. The classification process included determining the specialist areas of these applications such as educational and training, nursing, diagnosis and treatment, patient monitoring, testing and laboratories, and social networking. After the classification, a criterion to evaluate applications within individual categories, as well as in more general aspects such as their performance, security, user interface, and other software quality attributes, was developed. Test data was used to test the applications using the developed evaluation criteria, and the results were then used to determine the apps with surpassing features. As per the category-wise results, Medicine References and Education &amp; Training categories had applications that had better scores than other categories. However, an equally important finding concluded that there are not enough applications to help with lab testing, and this gap needs to be filled. Most applications lacked the usability aspect and needed work in user interface (UI) and user experience (UX) areas.

The easy-to-hard training advantage with real-world medical images

Many medical professions require practitioners to perform visual categorizations in domains such as radiology, dermatology, and neurology. However, acquiring visual expertise is tedious and time-consuming and the perceptual strategies mediating visual categorization skills are poorly understood. In this paper, the Ease algorithm was developed to predict an item’s categorization difficulty (Ease value) based on the item’s perceptual similarity to all within-category items versus between-category items in the dataset. In this study, Ease values were used to construct an easy-to-hard and hard-to-easy training schedule for teaching melanoma diagnosis. Whereas previous visual training studies suggest that an easy-to-hard schedule benefits learning outcomes, no studies to date have demonstrated the easy-to-hard advantage with complex, real-world images. In our study, 237 melanoma and benign images were collected for training and testing purposes. The diagnostic accuracy of images was verified by an expert dermatologist. Based on their Ease values, the items were grouped into easy, medium, and hard categories, each containing an equal number of melanoma and benign lesions. During training, participants categorized images of skin lesions as either benign or melanoma and were given corrective feedback after each trial. In the easy-to-hard training condition, participants learned to categorize all the easy items first, followed by the medium items, and finally the hard items. Participants in the hard-to-easy training condition learned items in the reverse order. Post-training results showed that training in both conditions transferred to the classification of new melanoma and benign images. Participants in the easy-to-hard condition showed modest advantages both in the acquisition and retention of the melanoma diagnosis skills, but neither scheduling condition exhibited a gross advantage. The Ease values of the items predicted categorization accuracy after, but not before training, suggesting that the Ease algorithm is a promising tool for optimizing medical training in visual categorization.

The effect of feedback delay on perceptual category learning and item memory: Further limits of multiple systems.

Delayed feedback during categorization training has been hypothesized to differentially affect 2 systems that underlie learning for rule-based (RB) or information-integration (II) structures. We tested an alternative possibility: that II learning requires more precise item representations than RB learning, and so is harmed more by a delay interval filled with a confusable mask. Experiments 1 and 2 examined the effect of feedback delay on memory for RB and II exemplars, both without and with concurrent categorization training. Without the training, II items were indeed more difficult to recognize than RB items, but there was no detectable effect of delay on item memory. In contrast, with concurrent categorization training, there were effects of both category structure and delayed feedback on item memory, which were related to corresponding changes in category learning. However, we did not observe the critical selective impact of delay on II classification performance that has been shown previously. Our own results were also confirmed in a follow-up study (Experiment 3) involving only categorization training. The selective influence of feedback delay on II learning appears to be contingent on the relative size of subgroups of high-performing participants, and in fact does not support that RB and II category learning are qualitatively different. We conclude that a key part of successfully solving perceptual categorization problems is developing more precise item representations, which can be impaired by delayed feedback during training. More important, the evidence for multiple systems of category learning is even weaker than previously proposed. (PsycINFO Database Record

Does categorization training change the encoding of face identity?

Does categorization training change the encoding of face identity?

Evaluating the Success of Project Management Information Systems’ Comprehensive Planning in Industrialized Building Systems

The main reason for the failure of most of management information systems is referred to the problems that arise from management factors and strategic components, expectations, complexity, and risks. Finally, the main applicable proposal is giving importance by managers and analysts to the fit between organization and its information system as well as giving attention to training categories in implementation of information system, preparing financial supports and human resource in design, implementation, and development of information system, strengthening progressive factors, and weakening restrictive factors in organization by management to design or develop the system and finally, preparing organizational infrastructures to implement information system. Introduction of an information system has a strong managerial, organizational, and technological impact on industrial building methods. One of the objectives of this study was to identify and introduce key success factors of Project Management Information Systems (PMIS) in Industrialized Building Systems (IBS). The results of hypotheses showed that all dimensions of PMIS influence the performance of projects. But in the second phase of the study, sub-factors were classified using hierarchical analysis approach. This classification helps managers to identify the most important factors and obtain better results concentrating on the main factors compared with using PMIS.

Features of the calculation of the plan of forming single-group trains in market conditions

The article is analyzing the possibility of using an additional parameter for the calculation of the formation of one-group trains, allowing to take into account the costs associated with the motion of rolling stock. A study was carried out at a number of railway sections of the Belarusian Railways aimed at identifying factors that affect the sectional speed of freight trains. The analysis of the obtained results made it possible to conclude that there is a relationship between the train category and its sectional speed. Depending on the traffic intensity and capacity of the site, the difference between the sectional speeds of transit and sectional trains in the investigated areas was from 2 to 18 %. In the development of the train formation plan, it is proposed to apply a new additional parameter for saving costs for cars and locomotives in transit trains Rn . Authors propose formulas for calculating the value of this parameter. The influence of some factors was analyzed including the train composition and mass, the length of the section, the power of the car stream on cost savings when using the proposed parameter in the calculations. The application of the parameter of cost savings in the cars and locomotives in transit trains makes it possible to justify the allocation of separate streams of car traffic for independent purposes, which would be considered economically unprofitable when developing a train formation plan based on calculations performed by traditional methods.

The introduction of intellectual system for evaluating professional abilities of applicants into the activities of educational institutions

The object of research is the methods and means of professional identification of applicants. The research aims to provide applicants with scientifically based decision support for choosing a field of study. The introduction of intelligent decision support systems in the process of self-actualization of applicants will significantly improve the effectiveness of vocational guidance activities of educational institutions. One of the biggest problems of the intellectualization of systems for assessing abilities and achievements is that the test results of applicants represent a lot of fuzzy data. At the same time, the correctness of data separation significantly depends on the construction of a fuzzy set of features for the conclusion of a diagnostic solution. In addition, the most common tests do not take into account the requirements for specialists of the construction industry. The basis of the developed system is tests to determine the personality structure of the Integrated Professional Orientation Diagnostics “Applicant”. This system contains reference information about the vocational category of training and tests for determining the structure of the individual. Conclusions are based on techniques that allow to predict the success of activities in various industries. The ability of a person to a certain professional activity reflects the ability to acquire special knowledge and skills in the learning process. That is why in the course of the study the Applicant tests were used. To improve the reliability of assessing the professional abilities of the applicant, it is proposed to use an intelligent system based on the Takagi-Sugeno-Kang fuzzy neural network. This choice is due to the fact that the Takagi-Sugeno-Kang network has a number of features that provide it with advantages in solving the problem of matching the abilities of an applicant to the possibility of acquiring knowledge and skills in a particular specialty. In particular, the ability of fuzzy neural networks to separate linearly inseparable data. This ensures the ability of the system to isolate the natural abilities of applicants from a mixture of data. Compared to other means, the Takagi-Sugeno-Kang network makes it possible to solve the problem of classifying a very large amount of data by a network of smaller dimension.

Percepción de la dificultad de la tarea en jóvenes jugadores de fútbol

Los objetivos del presente estudio son: a) valorar la percepción de dificultad que tienen niños en iniciación deportiva frente a unas tareas establecidas según nivel de dificultad percibido por el entrenador; y b) identificar los patrones técnico-tácticos que resultan más complicados en su aprendizaje para niños en iniciación deportiva. Para ello se diseñó un estudio de carácter cualitativo cuya muestra fueron jóvenes jugadores en iniciación deportiva (n=11), sobre los que se aplicó una serie de tareas que su formador había ordenado en base a su percepción de la dificultad. Los resultados más destacados muestran que: a) el desarrollo de ciertas tareas de carácter analítico que se presuponen más sencillas suponen mayores grados de complejidad para los jóvenes jugadores de fútbol; y b) la percepción que tienen los entrenadores sobre la dificultad que suponen los juegos reducidos es mayor que la dificultad que perciben los jóvenes jugadores sobre este tipo de tareas. Estos datos pueden servir a los entrenadores en categorías de formación a orientar la preparación y el desarrollo de las tareas y de manera más global a llevar una metodología de entrenamiento más acorde a los jugadores.