• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.836-839
• /
• 2019

여러 분야에서 각광받는 딥러닝은 학습시간이 오래 걸리고, 고가의 장비들이 요구된다. 이러한 이유로 저사양 머신들을 이용한 분산 러닝 시스템들이 연구되기 시작했다. 본 논문은 " PSO 알고리즘을 이용한 분산 딥러닝 시스템" 을 개선했고, 그 결과 개선한 시스템의 머신 개수가 1 대 일 때 정확도가 92.8%까지 향상되었고, 머신 개수가 10 대 일 때 정확도가 93.4%까지 향상되었다. 이를 기반으로 저사양의 머신들을 결합한 분산 러닝 시스템이 고가의 장비를 사용하지 않고도 좋은 결과를 얻을 수 있다는 것을 확인했다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.116-118
• /
• 2020

GPU 기술과 빅데이터의 성장에 힘입어 최근 딥러닝 기술은 괄목할만한 성장을 이루었고, 구글, 페이스북, 우버 등의 빅데이터를 보유한 업체들과 슈퍼컴퓨팅분야에서는 이러한 빅데이터를 빠른 시간 안에 학습하기 위해 분산 딥러닝 기술을 연구해오고 있다. 이러한 대규모 분산 딥러닝에서는 집합 통신, IO 부하 등이 주요 병목으로 알려져 있다. 본 연구에서는 분산 딥러닝에서 시도되고 있는 주요 All-Reduce 알고리즘들에 대해 누리온 시스템에서 성능평가를 수행하였고, 512노드 이상의 대규모에서는 2D-torus 알고리즘이 우수한 성능을 보였다.

• Journal of The Korean Association of Information Education
• /
• v.12 no.3
• /
• pp.275-282
• /
• 2008

If a learner should fail to complete previous courses before moving on to the next level will face difficulty keeping up with it. Such personal tuition for those having trouble coping with their class is an issue that needs urgent addressing, which cannot be a burden only to teachers, this study has been conducted to sought for a solution. This study has developed and put into application a web based analyzing system to assess the area of deficit in students followed by obliterating accumulated learning deficits to impart assistance for their study. The subject of the study comprised of ten students from a school where the researcher is on duty and the field of assessment with the analyzing system were numbers and calculations. As a result, we could find out its efficiency in elevating their capability and interest in learning, which was proven to be statistically significant using ANOVA.

• Journal of Korea Multimedia Society
• /
• v.16 no.10
• /
• pp.1230-1237
• /
• 2013

Cyber lectures were expected to replace the traditional classroom lecture, yet they were criticized due to the inefficiency. With educational technology supplemented, cyber lectures have improved their efficiency through the coherence of superior planning ability and improvement in computer and network technology. This paper's purpose is to understand the factors that improve educational efficiency of cyber lectures. Cyber lecture for IT subject for software practice is created to test the educational efficiency as lecture on theories will not distinctively show the lecture efficiency. The survey was conducted to students and statistical analysis was done on collected data to analyze the factors that influence educational efficiency of created lecture.

• Journal of Engineering Education Research
• /
• v.14 no.3
• /
• pp.9-14
• /
• 2011

This study was conducted to compare the learning styles and self-leadership between engineering college students with high and low team performance records. About 70% of students in high team performance group showed learning styles of converger and accommodator, whereas about 67% of students in low team performance group showed learning styles of accommodator and diverger. In regard to self-leadership, high team performance group showed higher level of self-leadership, especially self-observation, self-punishment, natural reward strategies, visualizing successful performance, self-talk, and evaluating beliefs and assumptions. It is recommended to provide the engineering students with the specialized training program to complement their learning styles and self-leadership strategies.

• Journal of the Korea Convergence Society
• /
• v.9 no.1
• /
• pp.21-29
• /
• 2018

The purpose of this study is to explore supporting strategies for teaching and learning based on students' effect recognition and continuous intention in college flipped learning. It was analyzed 426 data by multivariate analysis of variance (MANOVA) by examining student's effect recognition and continuous intention on 15 flipped learning classes of K-university in Chungnam. The characteristics of learners were male, senior students, students who knew flipped learning, students who did not have previous experience, and students who were learning video at anytime. As a teaching strategy, it was found that effect recognition and continuous intention were high in the supplementary deepening flipped learning class and natural science or engineering area. As a teaching and learning supporting strategies, First, the university should develop and operate flipped class learning strategy program for females and low-grade students. Second, it should support the development of good flipped learning design and operation model of instructor. Third, it should support the development of high quality online learning contents that students can learn from time to time. Fourth, it should support the strengthening of teaching competency to develop and operate flipped learning classes. This study can be used as basic data to support and spread the effective flipped learning classes of the university in the future.

• Journal of Korean Home Economics Education Association
• /
• v.28 no.2
• /
• pp.21-39
• /
• 2016

The purpose of this study was to investigate the differences in the school life satisfaction of middle school students according to self-directed learning capability and emotion regulation strategy. The subject of this study were 499 middle school students in Daegu. In order to analyze the data, two way ANOVA were employed for analysis and $Scheff{\acute{e}}$ test for post-hoc analysis. The main finding of this study were as follows; First, there were significant differences in the school life satisfaction of middle school students according to self-directed learning capability and behavioral emotion regulation strategy. Second, there were significant differences in the school life satisfaction of middle school students by cognitive emotion regulation strategy. But there were no significant differences in the school life satisfaction according to self-directed learning capability and cognitive emotion regulation strategy. Third, there were significant differences in the school life satisfaction of middle school students according to negative avoidant and emotion regulation strategy. However there were no significant differences in the school life satisfaction according to self-directed learning capability and negative avoidant emotion regulation strategy.

• Journal of Advanced Navigation Technology
• /
• v.18 no.5
• /
• pp.468-475
• /
• 2014

The recommend system is getting more difficult real time recommend by lager preference data set, computing power and recommend algorithm. For this reason, recommend system is proceeding actively one's studies toward distribute processing method of large preference data set. This paper studied distribute processing method of large preference data set using hadoop distribute processing platform and mahout machine learning library. The recommend algorithm is used Co-occurrence Matrix similar to item Collaborative Filtering. The Co-occurrence Matrix can do distribute processing by many node of hadoop cluster, and it needs many computation scale but can reduce computation scale by distribute processing. This paper has simplified distribute processing of co-occurrence matrix by changes over from four stage to three stage. As a result, this paper can reduce mapreduce job and can generate recommend file. And it has a fast processing speed, and reduce map output data.

• The Journal of Korean Association of Computer Education
• /
• v.14 no.1
• /
• pp.1-12
• /
• 2011

Educational Programming Language has been reported to expand thinking ability and to give help in creative problem solving ability by numerous researches. Researchers are verifying the educational effects of EPL by applying it to various area, but researches in effective application of EPL is yet incomplete. Thus, for effective application of EPL, this research has verified the project completing ability depending on studying style targeted to college senior students. As results of verification, first, the results showed significant differences in project completing abilities depending on information processing methods, and learners who preferred self-reflecting introspection showed high scores. Second, in learning style the divergers showed the highest scores. This research suggested the necessity of guidance and detailed planning of self-reflecting introspective activity in ideas that would be realized by learners through searching for factors that could enhance the degree of project completion in programming learning using EPL.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.2
• /
• pp.405-416
• /
• 2022

Current AI technology is improving the quality of life by using machine learning based on data. When using machine learning, transmitting distributed data and collecting it in one place goes through a de-identification process because there is a risk of privacy infringement. De-identification data causes information damage and omission, which degrades the performance of the machine learning process and complicates the preprocessing process. Accordingly, Google announced joint learning in 2016, a method of de-identifying data and learning without the process of collecting data into one server. This paper analyzed the effectiveness by comparing the difference between the learning performance of data that went through the de-identification process of K anonymity and differential privacy reproduction data using actual financial data. As a result of the experiment, the accuracy of original data learning was 79% for k=2, 76% for k=5, 52% for k=7, 50% for 𝜖=1, and 82% for 𝜖=0.1, and 86% for Federated learning.