• Journal of the Korea Institute of Military Science and Technology
• /
• v.24 no.6
• /
• pp.591-601
• /
• 2021

Recently, infrared object detection(IOD) has been extensively studied due to the rapid growth of deep neural networks(DNN). Adversarial attacks using imperceptible perturbation can dramatically deteriorate the performance of DNN. However, most adversarial attack works are focused on visible image recognition(VIR), and there are few methods for IOD. We propose deep learning-based adversarial attacks for IOD by expanding several state-of-the-art adversarial attacks for VIR. We effectively validate our claim through comprehensive experiments on two challenging IOD datasets, including FLIR and MSOD.

• Steel and Composite Structures
• /
• v.49 no.1
• /
• pp.109-123
• /
• 2023

Headed studs welded to steel beams and embedded within the concrete of deck slabs are vital components of modern composite floor systems, where safety and economy depend on the accurate predictions of the stud shear resistance. The multitude of existing deck profiles and the complex behavior of studs in deck slab ribs makes developing accurate and reliable mechanical or empirical design models challenging. The paper addresses this issue by presenting a machine learning (ML) model developed from the natural gradient boosting (NGBoost) algorithm capable of producing probabilistic predictions and a database of 464 push-out tests, which is considerably larger than the databases used for developing existing design models. The proposed model outperforms models based on other ML algorithms and existing descriptive equations, including those in EC4 and AISC 360, while offering probabilistic predictions unavailable from other models and producing higher shear resistances for many cases. The present study also showed that the stud shear resistance is insensitive to the concrete elastic modulus, stud welding type, location of slab reinforcement, and other parameters considered important by existing models. The NGBoost model was interpreted by evaluating the feature importance and dependence determined with the SHapley Additive exPlanations (SHAP) method. The model was calibrated via reliability analyses in accordance with the Eurocodes to ensure that its predictions meet the required reliability level and facilitate its use in design. An interactive open-source web application was created and deployed to the cloud to allow for convenient and rapid stud shear resistance predictions with the developed model.

• Journal of the Korean association of regional geographers
• /
• v.12 no.2
• /
• pp.281-290
• /
• 2006

Map uses graphic language of dot, line and area to represent surface of the earth. Map has been adopted as tools for regional and cartography learning to improve graphicacy in geography education. Due to the rapid development in GIS and internet, practical use of maps has been extended in various study area. This Study is to develope web-based leaning materials for self-controled geography instruction. As learning materials for this aim, it has been constructed WebGIS for topography and thematic maps with boundary map of Chungbuk, digital map of Jochiwon(1:25,000), statistic data and field images. Function of WebGIS intend to improve skills on geo-information collection and spatial query, regional difference of spatial distribution. Individual learning using internet can make an improvement of learner centeredness and problem-solving. Finally, it will be expected to be suggest one of the education guide as blueprint in info-society.

• The Journal of the Korea Contents Association
• /
• v.15 no.2
• /
• pp.507-519
• /
• 2015

Along with the rapid diffusion of smart devices, smart learning has been taking place as a main pedagogy in education. Under these drastic changing circumstances, social presence and interaction between learners have been highlighted as key factors in educational research. In this light of rising importance, this study examined the effects of modality and smart device on users' perceived social presence and interaction experience in a smart learning environment. It conducted 2(text based interface vs. audio/video based interface) by 2(smartphone vs. tablet PC) between-subjects experiment. 80 participants were systematically recruited and randomly assigned to four conditions. The findings showed that social presence was higher in audio/video based interface condition than in text based interface condition. Modality also had a positive effect on learner's interaction experience. On the other hand, the effect of smart device is found to be statistically insignificant. Instead, interaction effect existed between modality and device on social presence. The result of this study suggests that the modality and characteristics of device should be considered seriously when designing interface of smart learning contents. The findings in this study provide future studies with heuristic implications by highlighting users' perceived cognition and experience.

• Journal of Convergence for Information Technology
• /
• v.10 no.11
• /
• pp.195-201
• /
• 2020

There have been many developments and innovations in the educational environments in line with the rapidly evolving information age. E-Learning is a representative example of this rapid evolution. However, E-Learning is challenging to maintain students' concentration because of the low engagement level and limited interactions between instructors and students. Additionally, instructors have limitations in identifying learners' concentration. This paper proposes a system that can measure E-learning users' concentration levels by detecting the users' eyelid movement and the top of the head. The system recognizes the eyelid and the top of the head and measures the learners' concentration level. Detection of the eyelid and the top of the head triggers an event to assess the learners' concentration level based on the users' response. After this process, the system provides a normalized concentration score to the instructor. Experiments with experimental groups and control groups were conducted to verify and validate the system, and the concentration score showed more than 90% accuracy.

• Korean Journal of Social Welfare
• /
• v.45
• /
• pp.341-373
• /
• 2001

This study seeks to identify the reform of vocational education plans so as to bring about a seamless transition from school to work. It puts forward a number of suggestions based upon an analysis of vocational education policies in Korean and U.S. schools, concerning the government's role, partnerships between education and industry, the educational system, curriculum and work-based learning. First, not only government initiatives but also close partnerships between education and industry are essential to help vocational education in school the transition to employment. Education and industry should work closely together to standardize certificate related skills and to have these skills reflected in the curriculum. Also the government should strive to provide guidelines for work-based learning and formulate standards for supervision and evaluation. Second, to facilitate the school to work transition, comprehensive schools should be promoted so that students have access to a greater ranger of vocational education. At the same time, an assessment system that certifies a mastering of the basic skills of those who undergo the education should be introduced, and it should be related to earn these certificates. Third, standardized vocational skills should be included in the curriculum so that students can acquire skills that are useful for industry. All the students in vocational and general high schools should have access both to general education, the foundation for lifelong learning and for employ ability, and to basic occupational skills which empower students in dealing with rapid changes of technology. Also a range of specialized vocational curricula should be offered so that students can opt for more specialized occupations; and they can select careers appropriate to their capability. Fourth, so that all students to have the opportunity to take part in work-based education, which is closely related to employment, various work-based learning programs should be offered to meet the needs of students and their educational conditions. Companies should for their part train students thoroughly in accordance with the standards of work-based education. In addition, supervisors should be stationed both in schools and companies in order to administer the students' work-based learning.

• Journal of Korean Society for Quality Management
• /
• v.49 no.4
• /
• pp.581-594
• /
• 2021

Purpose: The purpose of this study is to compare machine learning models for anomaly detection of mechanical facility equipment and suggest an anomaly detection system for mechanical facility equipment in subway stations. It helps to predict failures and plan the maintenance of facility. Ultimately it aims to improve the quality of facility equipment. Methods: The data collected from Daejeon Metropolitan Rapid Transit Corporation was used in this experiment. The experiment was performed using Python, Scikit-learn, tensorflow 2.0 for preprocessing and machine learning. Also it was conducted in two failure states of the equipment. We compared and analyzed five unsupervised machine learning models focused on model Long Short-Term Memory Variational Autoencoder(LSTM-VAE). Results: In both experiments, change in vibration and current data was observed when there is a defect. When the rotating body failure was happened, the magnitude of vibration has increased but current has decreased. In situation of axis alignment failure, both of vibration and current have increased. In addition, model LSTM-VAE showed superior accuracy than the other four base-line models. Conclusion: According to the results, model LSTM-VAE showed outstanding performance with more than 97% of accuracy in the experiments. Thus, the quality of mechanical facility equipment will be improved if the proposed anomaly detection system is established with this model used.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.2
• /
• pp.171-177
• /
• 2021

Through the rapid developments in machine learning, there have been diverse utilization approaches not only in industrial fields but also in daily life. Implementations of machine learning on financial data, also have been of interest. Herein, we employ machine learning algorithms to store sales data and present future applications for fintech enterprises. We utilize diverse missing data processing methods to handle missing data and apply gradient boosting machine learning algorithms; XGBoost, LightGBM, CatBoost to predict the future revenue of individual stores. As a result, we found that using median imputation onto missing data with the appliance of the xgboost algorithm has the best accuracy. By employing the proposed method, fintech enterprises and customers can attain benefits. Stores can benefit by receiving financial assistance beforehand from fintech companies, while these corporations can benefit by offering financial support to these stores with low risk.

• Journal of the Korea Institute of Building Construction
• /
• v.21 no.1
• /
• pp.71-78
• /
• 2021

The spread of COVID-19 and the advent of the Fourth Industrial Revolution have significantly affected the nature of college education causing many changes to the way it is conducted. One of these changes is the expansion of online education. The purpose of this study was to analyze the difficulties experienced by learners due to the transition to rapidly expanding online education at domestic universities, and to seek ways to support their learning through this new online platform. Results of a questionnaire showed that learners experienced difficulties in their interactions with professors because of the rapid transition to online education without adequate preparation. It was determined that there were not enough opportunities for communication between learners and professors as a result of non-face-to-face online education, and that learners did not receive Q&A or feedback quickly enough. The study also examined ways to ways to improve the effectiveness of online learning. Students showed a high preference for items such as "appropriate guidance regarding announcements such as lecture schedules," "providing lecture notes as learning materials."

• Journal of Information Science Theory and Practice
• /
• v.10 no.spc
• /
• pp.143-153
• /
• 2022

With the development of networks and the increase in the number of network devices, the number of cyber attacks targeting them is also increasing. Since these cyber-attacks aim to steal important information and destroy systems, it is necessary to minimize social and economic damage through early detection and rapid response. Many studies using machine learning (ML) and artificial intelligence (AI) have been conducted, among which payload learning is one of the most intuitive and effective methods to detect malicious behavior. In this study, we propose a preprocessing method to maximize the performance of the model when learning the payload in term units. The proposed method constructs a high-quality learning data set by eliminating unnecessary noise (stopwords) and preserving important features in consideration of the machine language and natural language characteristics of the packet payload. Our method consists of three steps: Preserving significant special characters, Generating a stopword list, and Class label refinement. By processing packets of various and complex structures based on these three processes, it is possible to make high-quality training data that can be helpful to build high-performance ML/AI models for security monitoring. We prove the effectiveness of the proposed method by comparing the performance of the AI model to which the proposed method is applied and not. Forthermore, by evaluating the performance of the AI model applied proposed method in the real-world Security Operating Center (SOC) environment with live network traffic, we demonstrate the applicability of the our method to the real environment.