• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.317-318
• /
• 2022

본 논문에서는 클러스터 기반 엣지 모델에서 자원을 효율적으로 사용할 수 있는 지능형 엣지 자원 할당 정책 모델을 제안한다. 최근 쿠버네틱스 기반 클러스터 엣지 시스템 개발 연구 다양한 방향에서 진행 중이다. 따라서, 본 논문에서는 클러스터-엣지 모델 구조를 소개하고 이 모델에서 컴퓨팅 자원을 가진 워커에 컴퓨팅 오프로딩 서비스를 효율적으로 사용할 수 있는 최적의 지능형 클러스터-엣지 컴퓨팅 자원 정책을 생성하는 구조 및 알고리즘을 제안한다.

• Journal of The Korean Association For Science Education
• /
• v.35 no.4
• /
• pp.751-772
• /
• 2015

Valid and effective science education would require research-based decisions on multiple aspects of science education including policy decisions, science curriculum development, designing teaching resources and methods. However, this has not been the case. In order to provide a research base for science education practices and policy-making, this study reviewed research articles published in major science education research journals in South Korea in the last ten years. The analysis was focused on 8 areas including student conceptions, student thinking, inquiry, affective domain, student ideas about science, science curriculum, students' learning and classroom activity, and student learning in informal settings. General research trends found include: First, science education research conducted for the past decade focused on a certain limited topics/areas. Second, research participants were also limited to certain grade levels or types of students. Third, rather than examining developmental processes descriptive research was prevalent. Fourth, there was a lack of research on developing new areas of study or research on generation of new perspectives, theories or tools. Fifth, many studies were related to school science learning while relatively less studies were about other areas that would impact students' future. Based on the results, we suggest several implications for science curriculum development, policy development, science teaching and learning resources, and others.

• Journal of Practical Engineering Education
• /
• v.14 no.1
• /
• pp.219-224
• /
• 2022

Q-Learning is a technique widely used as a basic algorithm for reinforcement learning. Q-Learning trains the agent in the direction of maximizing the reward through the greedy action that selects the largest value among the rewards of the actions that can be taken in the current state. In this paper, we studied a policy that can speed up agent training using Q-Learning in Frozen Lake 8×8 grid environment. In addition, the training results of the existing algorithm of Q-learning and the algorithm that gave the attribute 'direction' to agent movement were compared. As a result, it was analyzed that the Q-Learning policy proposed in this paper can significantly increase both the accuracy and training speed compared to the general algorithm.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.3
• /
• pp.9-17
• /
• 2021

In this paper, an autonomous multi-agent path planning using reinforcement learning for monitoring of infrastructures and resources in a computationally distributed system was proposed. Reinforcement-learning-based multi-agent exploratory system in a distributed node enable to evaluate a cumulative reward every action and to provide the optimized knowledge for next available action repeatedly by learning process according to a learning policy. Here, the proposed methods were presented by (a) approach of dynamics-based motion constraints multi-agent path-planning to reduce smaller agent steps toward the given destination(goal), where these agents are able to geographically explore on the environment with initial random-trials versus optimal-trials, (b) approach using agent sub-goal selection to provide more efficient agent exploration(path-planning) to reach the final destination(goal), and (c) approach of reinforcement learning schemes by using the proposed autonomous and asynchronous triggering of agent exploratory phases.

• Journal of Digital Convergence
• /
• v.16 no.3
• /
• pp.527-532
• /
• 2018

This study aims to examine the effectiveness of leisure activity on learning agility and to verify the mediative effectiveness of interpersonal relationship between leisure activity and learning agility for youths. To achieve these purpose, 6,637 data that sampled by 'IEA ICCS 2016' were analyzed with structural equation. The results of study were as follow. First, leisure activity directly effects on learning agility for youths. Second, leisure activity directly effects on interpersonal relationship. Third, interpersonal relationship directly effects on learning agility and partial mediative effects between leisure activity and learning agility. These results implies that youths' leisure activity is not only directly affecting learning agility, but also affecting interpersonal relationship. And this interpersonal relationship also effects learning agility. Based on these results, some practical implications were proposed to develop learning agility in terms of leisure activity and interpersonal relationship.

• Journal of Digital Convergence
• /
• v.13 no.11
• /
• pp.25-33
• /
• 2015

In Korea, despite the increase in demand for Bigdata manpower, institutions and universities to educate and train Bigdata manpower are not yet much. Therefore, this study investigated the status regarding the recognition on Bigdata of universities students and presented a direction for educating Bigdata manpower at the university. In order to accomplish this purpose, this study surveyed and analyzed the students' understanding of Bigdata, the awareness of the students about the social impact of Bigdata, the learning intention of the students on Bigdata and presented Implications for Bigdata workforce development. As a result, despite of the somewhat difference in understanding for the Bigdata, it was found that their awareness about the impact of Bigdata is very positive. And this study showed the need of universities' and government' political effort for Bigdata workforce development, because it was investigated that students' intentions of learning for Bigdata is proportional to students' understanding levels and learning experience for Bigdata.

• Journal of Digital Convergence
• /
• v.16 no.2
• /
• pp.83-91
• /
• 2018

The purpose of this study was to examine teacher's emotional leadership practices and to suggest their educational policy implication. To do so, a survey was conducted on Chinese-Korean students in Yanbian area. Data samples were 605 students at three high schools. The major results are summarized as follows: First, the teacher's emotional leadership is an important variable to improve student's self-directed learning. Second, it is suggestive that the teacher should focus on developing students personal competences such as self awareness and self management rather than social competences. Lastly, the teacher should focus on and improve satisfaction of student's school life first, and try to increase student's self-directed learning. However, various variables that may influence teacher's emotional leadership should be included in the further study.

• 한국디지털정책학회:학술대회논문집
• /
• 2005.06a
• /
• pp.559-570
• /
• 2005

This paper presents how to establish an efficient e-Learning environment using open source software. A LMS with additional functionalities on the top of dotLRN. which is a open source project for LMS, is presented. Additional functionalities include modification of the language for Korean, adoption of SCORM educational standard, and management of learning outcome. This system had been serviced for Kongju cyber university for one year on stable basis. The scope of this paper covers introduction, characteristics review, and the learner's learning outcome analysis.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.9
• /
• pp.213-221
• /
• 2021

In this study, an simulation app was developed as one of the methods to help learners better understand the eight process scheduling policies of multi-programming. In learning, an app in the form of a simulation should provide a realistic environment and allow learners to practice. To do this, the needs of the learners were investigated and analyzed, and the purpose was set, designed, and programmed based on the learners' understanding. And it was shown that the apps as a tool to simulate the created eight scheduling policies are performing well. In particular, it was shown that the problem of not having a step-by-step various diagram and explanation for step-by-step various inputs, which is a limitation of paper textbooks, can be solved using these tools.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.644-649
• /
• 2021

Self-Imitation Learning is a simple off-policy actor-critic algorithm that makes an agent find an optimal policy by using past good experiences. In case that Self-Imitation Learning is combined with reinforcement learning algorithms that have actor-critic architecture, it shows performance improvement in various game environments. However, its applications are limited to reinforcement learning algorithms that have actor-critic architecture. In this paper, we propose a method of applying Self-Imitation Learning to Deep Q-Network which is a value-based deep reinforcement learning algorithm and train it in various game environments. We also show that Self-Imitation Learning can be applied to Deep Q-Network to improve the performance of Deep Q-Network by comparing the proposed algorithm and ordinary Deep Q-Network training results.