training problem
Recently Published Documents


TOTAL DOCUMENTS

105
(FIVE YEARS 31)

H-INDEX

12
(FIVE YEARS 0)

Author(s):  
Jie Yang ◽  
Lian Tang ◽  
Xin-Wei Li

With the application of artificial intelligence in many social fields, the research of human behavior recognition and non-contact detection of human physiological parameters based on face recognition and other technologies has developed rapidly, and the application of artificial intelligence in culture, sports and entertainment has also begun to rise. How to apply the existing mature technology to the sports intelligence training system taking table tennis as an example is a hot issue worthy of study. In this paper, a comprehensive intelligent table tennis training system and platform based on Convolutional Neural Network face recognition and face heart rate detection is designed, which is mainly used to solve the philosophical training problem in table tennis. In the system place, an identification cameras is set at the entrance of table tennis training places, which is used for table tennis players’ sign-in and training table number allocation, and an intelligent analysis cameras is set above each intelligent training table, which is used for detecting the face and heart rate of table tennis players. Each intelligent training platform consists of intelligent voice control unit, server, camera, industrial control computer, monitor and other terminal modules. The member data center constitutes the platform of intelligent table tennis training system.


Author(s):  
Jaime A. Fernández ◽  
Ernesto Baena

In dual disorder, two serious and chronic disorders converge that are still a challenge to health and social care networks. In this context, families play an important role in keeping these people included in the community. Dual disorder is associated with a series of negative effects on the family environment, with a greater burden of care and conflict. For this article, four models of family intervention in dual disorder have been reviewed. Conclusions. Family intervention has proven to be an important element of dual disorder treatment. The four intervention programs presented coincide in share some common components: single / multi-family intervention, theoretical bases of the models of with proven efficacy, psychoeducation, communication training, problem solving, and the motivational interview across the entire program. Even so, some areas still persist without improvements and areas that do not improve persist and the results are not conclusive, so it is necessary to continue looking for formulas that point towards more flexible therapeutic resources according to the needs and circumstances of each of these people.


Author(s):  
Yangyang Zhao ◽  
Zhenliang Ma ◽  
Xinguo Jiang ◽  
Haris N. Koutsopoulos

Unplanned events present significant challenges for operations and management in metro systems. Short-term ridership prediction can help agencies to better design contingency strategies under unplanned events. Though many short-term prediction methods have been proposed in the literature, most studies focused on typical situations or planned events. The study develops methods for the short-term metro ridership prediction under unplanned events. It explores event impact representation mechanisms and deals with the imbalanced data training problem in building the prediction model under unplanned events. Typical machine learning and deep learning methods are developed for exploration. A large-scale automatic fare collection (AFC) dataset and event record data for a heavily used metro system are used for empirical studies. The analysis found that the same type of unplanned event shares a similar and consistent demand change pattern (with respect to the demand under typical situations) at the station level. The synthetic minority oversampling technique (SMOTE) can enrich the ridership observations under unplanned events and generate a balanced dataset for model training. Given the occurrence of unplanned events, the results show that a combination of demand change ratio and the SMOTE oversampling technique enables the prediction models to learn the impact of unplanned events and improve the prediction accuracy under unplanned events. However, the oversampling methods (i.e., SMOTE and replication) slightly deteriorate the prediction accuracy for ridership under normal conditions. The findings provide insights into mechanisms for disruption impact representation and oversampling imbalanced data in model training, and guide the development of models for short-term prediction under unplanned events.


2021 ◽  
Vol 25 (4) ◽  
pp. 23-38
Author(s):  
T. A. Kustitskaya ◽  
R. V. Esin

The aim of the study. The fourth industrial revolution demands highly qualified personnel as important factor of economic growth, which imposes serious requirements on the formation of key and subject competencies among graduates of higher educational institutions. A particularly important role is assigned to the mathematical competence which is required to solve complex and science-intensive problems. Given the growing share of e-learning and distance learning at the university, it is necessary to intensively develop the methodology for mathematical competence formation in the electronic environment, and create effective teaching tools on its basis. The current level of digitalization of education already allows organizing independent work of students in the electronic environment at a sufficiently high level. In the literature we can find various methods and tools, aimed at the formation of the cognitive component of competencies. However, the issue of skills’ development in the electronic environment is still underrepresented. The purpose of this study is to develop a methodology for creating electronic training problems, which aims at forming a practical component of mathematical competence – the competency of solving mathematical problems.Materials and methods. In the study we performed a comparative analysis of scientific and methodological literature, regulatory and methodological documents, as well as professional and federal educational standards of higher education. The development of a model of electronic training problems was carried out using methods of structural modeling. The developed methodology was implemented in the educational process, and the confirmation of its effectiveness was obtained by statistical analysis of the results of the pedagogical experiment.Results. We proposed a methodology for electronic training problems development aimed at formation of mathematical problems solving competency. The methodology is based on existing approaches to problem solving formalization. In the presented structural model of an electronic training problem, the aspects of problem solving discovered earlier by other authors, are supplemented by the contextual aspect. This aspect is intended for linking the regarded problem with the material, studied at the moment and, if possible, with future professional activity of a student. The proposed methodology for organizing feedback in an electronic training problem contributes to the formation of metacognitive skills among students through the elements of tutoring.Conclusion. On the basis of the proposed methodology, 8 electronic training problems were developed for the course “Probability and Mathematical Statistics” and tested in the educational process of the Siberian Federal University. The effectiveness of the electronic training problems for the development of mathematical problems solving competency was assessed in the course of a pedagogical experiment. The purpose of the experiment was to study the impact of the electronic training problems in the competency formation for particular topics of the course. Using student’s test for independent samples and the Mann-Whitney test we confirmed that the designed electronic training problems positively affect the formation of mathematical problems solving competency. In the future, the proposed methodology can be included in the teaching toolkit for the formation of mathematical competence in an electronic environment.


2021 ◽  
Vol 12 (4) ◽  
pp. 209-215
Author(s):  
D. I. Chitalov ◽  

The research, the results of which are presented in this article, is devoted to the development of a software module with a graphical user interface that provides a modification of the computational mesh based on the dsmcInitialise utility, which is used at the preprocessing stage of numerical modeling of continuum mechanics problems using the OpenFOAM software environment. The paper describes the existing graphical shells for working with OpenFOAM with an indication of their shortcomings, formulates the relevance of the work, and defines the goals and objectives of the study. The article presents the features of the direct Monte Carlo simulation method, a description of the dsmcInitialise utility integrated into OpenFOAM and designed for such modeling, as well as a description of the corresponding dictionary file with parameters. The article includes diagrams of the structure and logic of the application, describes the technology stack used. The results of the application of the program on the example of one of the training problem of OpenFOAM are presented. The final conclusions are formulated, as well as the provisions that determine the scientific novelty of the research, and its intended practical value is determined. A link to the repository with the source code of the presented software module is provided.


2021 ◽  
Vol 3 (1) ◽  
pp. 19-28
Author(s):  
Arini Rosa Sinensis ◽  
Thoha Firdaus ◽  
T Hardila ◽  
Nopitasari Nopitasari ◽  
N Saiputri

The challenge of 21st century education is to prepare human resources who are required to have skills, one of which is problem solving. The Problem Based Learning (PBL) learning model is considered effective in training problem-solving skills because it has constructivist characteristics with science learning. The purpose of this study was to analyze students' problem solving skills in the Light material through PBL learning. The research method used is quantitative with descriptive analysis. The research sample was 22 Tanah Merah Integrated Junior High School students grade IX. The data collection technique used the observation sheet of problem solving skills. The results showed that there was an increase in students' problem solving skills from the first to the third experiment. The increase with the highest percentage was in the third experiment with indicators of problem solving/ investigation by 90.9%. The average data for each problem solving indicator shows that students can understand the problem by 80.78%, collect data by 69.63%, carry out problem solving/investigation by 78.46%, and make conclusions by 66.67%. These results indicate that the Problem Based Learning model can be used as a science learning construction in improving problem solving skills.


This paper investigates multilevel initializa- tion strategies for training very deep neural networks with a layer-parallel multigrid solver. The scheme is based on a continuous interpretation of the training problem as an optimal control problem, in which neu- ral networks are represented as discretizations of time- dependent ordinary differential equations. A key goal is to develop a method able to intelligently initialize the network parameters for the very deep networks en- abled by scalable layer-parallel training. To do this, we apply a uniform refinement strategy across the time domain, that is equivalent to refining in the layer di- mension. This refinement algorithm builds good ini- tializations for deep networks with network parameters coming from the coarser trained networks. The effec- tiveness of multilevel strategies (called nested iteration) for training is investigated using the Peaks and Indian Pines classification data sets. In both cases, the vali- dation accuracy achieved by nested iteration is higher than non-nested training. Moreover, run time to achieve the same validation accuracy is reduced. For instance, the Indian Pines example takes around 25% less time to train with the nested iteration algorithm. Finally, using the Peaks problem, we present preliminary anec- dotal evidence that the initialization strategy provides a regularizing effect on the training process, reducing sensitivity to hyperparameters and randomness in ini- tial network parameters.


2021 ◽  
Vol 336 ◽  
pp. 08012
Author(s):  
Lu Liu ◽  
Guobao Feng

In this paper, a new architecture of three-dimensional deep convolutional generative adversarial network(3D-DCGAN) is specially defined to solve the unstable training problem of GAN and make full use of the information involved in polarimetric data. Firstly, a data cube with nine components of polarimetric coherency matrix are directly used as the input features of DCGAN. After that, a 3D convolutional model is designed as the components of generator and discriminator to construct the 3D-DCGAN, which considers the effective feature extraction capability of 3D convolutional neural network(CNN). Finally parameters of the network are fine-tuned to realize the polarimetric SAR image classification. The experiments results show the feasibility and efficiency of the proposed method.


Sign in / Sign up

Export Citation Format

Share Document