• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.3
• /
• pp.147-152
• /
• 2023

This paper is a study of a linear algebra course taught in a high school on-demand course. Compared to the regular course, flipped learning was added to the course, and applications to signal processing related problems were covered in consideration of students' career aspirations. Overall, the class was a mixture of traditional lectures and flipped learning. Flipped learning was implemented twice. The flipped class consisted of pre-class, in-class and post-class. To verify the effectiveness of the course, a survey was conducted and most of the evaluation items were above 4. The topics of the flipped learning were Markov chains and least squares problem, which are very important in the field of signal processing.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.05a
• /
• pp.359-365
• /
• 2008

CP(Counterpropagation)알고리즘은 서로 다른 두 개의 신경망이 하나로 결합 된 혼합형 모델로서, 다른 신경망 모델에 비해 비교적 단순하고 빠른 학습 속도를 보인다. 그러나 CP 알고리즘은 다양한 패턴이 입력되면 충분한 경쟁층의 수가 설정되지 않아 학습이 불안정하고, 출력층에서 연결강도를 조정할 때 일반적인 학습률 조정방법으로 불안정한 학습 결과를 보인다. 이러한 문제점을 해결하기 위해 다수의 경쟁층을 설정하여 경쟁층에서 패턴 분류의 정확성을 높이고, 입력 벡터와 승자 뉴런의 대표 벡터간의 차이와 승자 빈도수를 반영하여 학습률을 동적으로 조정하여 경쟁층에서의 학습이 안정적으로 진행되도록 하고, 출력층에서 연결강도를 조정할 때 모멘텀(momentum)학습법을 적용한 개선된 CP 알고리즘이 제안되었다. 본 논문에서는 개선된 CP 알고리즘에서 경쟁층의 수를 효율적으로 설정하기 위해 퍼지 제어 기법을 이용하여 경쟁층의 수를 결정하는 방법을 제안한다. 제안된 방법은 CP 알고리즘에 입력되는 패턴의 정보를 이용하여 퍼지 소속 함수를 설계하고 입력에 대한 소속도를 계산한 후, 퍼지 제어 규칙을 적용하고, Mamdani의 Min_Max 추론 방법으로 추론한다. 퍼지 추론을 통해 최종적으로 얻어진 값을 무게 중심법으로 비퍼지화 하여 최종적으로 개선된 CP 알고리즘의 경쟁층의 수를 결정하는데 적용한다. 제안된 방법의 학습 및 인식 성능을 평가하기 위해, 숫자, 영어 등과 같이 다양한 패턴을 실험에 적용한 결과, 제안된 방법이 경쟁층의 수를 결정하는데 효과적임을 확인할 수 있었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.12
• /
• pp.2622-2627
• /
• 2010

The accuracy of face recognition used unmanned security system is very important and necessary. However, face recognition is a lot of restriction due to the change of distortion of face image, illumination, face size, face expression, round image. We propose a hybrid neural network for improve the performance of the face recognition. The proposed method is consisted of SOM and LVQ. In order to verify usefulness of the proposed method, we make a comparison between eigenface method, hidden Markov model method, multi-layer neural network.

• Journal of The Korean Association of Information Education
• /
• v.15 no.1
• /
• pp.77-84
• /
• 2011

Distance learning makes it possible for educational consumers to study anywhere or anytime. In order to provide effective distance learning, it requires not only educational process and educational results to be closely correlated but also a method for teachers and students to communicate immediately. We devised a dialogue system like a chatting program containing an editor. With this program, students can learn a programming practice by chatting between a teacher and a student. We also incorporated a source code visualization technique based on diff algorithm so that the visualization technique makes it possible for students to recognize the remedies by teachers visually. The result comes by mixing the merits of a dialogue system giving a close interrelationship among users into a programming language education. We expect that the successful result would be tantamount to a face-to-face education.

• Journal of The Korean Association of Information Education
• /
• v.25 no.4
• /
• pp.653-664
• /
• 2021

This study proposes a STEAM (Science, Technology, Engineering, Arts, and Mathematics) program for career exploration of middle school students. The proposed program utilizes VR (Virtual Reality) for new digital technology and webtoon as a popular cultural element. It enables the students to investigate promising fields and experience them virtually for themselves. We design the program based on the 2015 revised curriculum, which enhances the learning effects with existing subjects. In particular, the program provides a hybrid education to combine contact and untact classes considering the COVID-19 situation. The educational goal of the proposed program is to improve creativity and convergence capability. Specifically, it aims to prepare an educational foundation that integrates new digital technologies into education and applies the programs to school education fields. To prove the effectiveness of the developed program, we applied the program to the second graders of A middle school located in Seoul. We expect that the proposed program helps students learn how to utilize new digital technologies and explore future career paths.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.531-534
• /
• 2022

신경망 기반 심층학습 기술은 대화처리 분야에서 대폭적인 성능 개선을 가져왔다. 특히 GPT-2와 같은 대규모 사전학습 언어모델을 백본 네트워크로 하고 특정 도메인 타스크 대화 데이터에 대해서 미세조정 방식으로 생성되는 종단형 대화모델의 경우, 해당 도메인 타스크에 대해서 높은 성능을 내고 있다. 하지만 이런 연구들은 대부분 하나의 도메인에 대해서만 초점을 맞출 뿐 싱글 모델로 두 개 이상의 도메인을 고려하고 있지는 않다. 특히 순차적인 미세 조정은 이전에 학습된 도메인에 대해서는 catastrophic forgetting 문제를 발생시킴으로써 해당 도메인 타스크에 대한 성능 하락이 불가피하다. 본 논문에서는 이러한 문제를 해결하기 위하여 MultiWoz 목적지향 대화 데이터에 오픈 도메인 칫챗 대화턴을 유사도에 기반하여 추가하는 데이터 증강 방식을 통해 사용자 입력 및 문맥에 따라 MultiWoz 목적지향 대화와 오픈 도메인 칫챗 대화를 함께 생성할 수 있도록 하였다. 또한 목적지향 대화와 오픈 도메인 칫챗 대화가 혼합된 대화에서의 시스템 응답 생성 성능을 평가하기 위하여 오픈 도메인 칫챗 대화턴을 수작업으로 추가한 확장된 MultiWoz 평가셋을 구축하였다.

• The KIPS Transactions:PartA
• /
• v.11A no.5
• /
• pp.407-412
• /
• 2004

The e-learning system uses multimedia content for efficiency of a education and learner's convenience. With the web based e-learning, various multimedia services are difficult on account of physical factors such as network state and system performance. Also on the off line based e-learning, we have defects an insufficiency of interactions among learners and instructors, and an insufficiency of an adaptation about the latest information. In this study, we designed a multi-tier e-learning system using hybrid multimedia method, The proposed system can provide interaction and adaptability such as web based systems, and it can provide various multimedia content such as off line based systems. In experiments results, we know the proposed system improved users' convenience and an adaptability about the latest information, and interactions among learners and instructors when compare with existing e-learning system.

• Journal of Practical Engineering Education
• /
• v.15 no.1
• /
• pp.183-192
• /
• 2023

This study analyzes the performance of flipped learning, an offline class method conducted in a degree-linked work-learning parallel system. Training in the work-study parallel system, which is conducted as part of job competency improvement, has thoroughly adhered to the offline method, but in line with COVID-19, unlike before, it is changing in the direction of using the online method more actively. However, educational methods such as flipped learning are not new because the degree-linked operation is applied to the academic system and education method of universities and is practically the same form as general university education. Therefore, it is necessary to analyze the educational performance and complementary points of flipped learning, which has recently been expanded in the degree-linked work-study parallel system, considering the characteristics of this system, in which classes are held only on weekends. As a result of statistical analysis based on the survey, some of the outcomes of flipped learning have been confirmed, and in order to increase the performances, it is necessary to continuously seek out specific measures to encourage learning and expand communication between instructors and students.

• Journal of Intelligence and Information Systems
• /
• v.12 no.3
• /
• pp.67-82
• /
• 2006

We propose the HBEM(Hybrid Behavior Evolution Model) composed of rule classification and evolutionary neural network using rule descriptor and link descriptor for evolutionary behavior of virtual robots. In our model, two levels of the knowledge of behaviors were represented. In the upper level, the representation was improved using rule and link descriptors together. And then in the lower level, behavior knowledge was represented in form of bit string and learned adapting their chromosomes by the genetic operators. A virtual robot was composed by the learned chromosome which had the best fitness. The composed virtual robot perceives the surrounding situations and they were classifying the pattern through rules and processing the result in neural network and behaving. To evaluate our proposed model, we developed HBES(Hybrid Behavior Evolution System) and adapted the problem of gathering food of the virtual robots. In the results of testing our system, the learning time was fewer than the evolution neural network of the condition which was same. And then, to evaluate the effect improving the fitness by the rules we respectively measured the fitness adapted or not about the chromosomes where the learning was completed. In the results of evaluating, if the rules were not adapted the fitness was lowered. It showed that our proposed model was better in the learning performance and more regular than the evolutionary neural network in the behavior evolution of the virtual robots.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10a
• /
• pp.139-142
• /
• 2006

본 논문에서는 햅틱 피드백과 스테레오 비쥬얼 큐를 혼합한 다중 감각 기반의 지능형 3차원 형상 분석 방법을 소개한다. 지능형 형상 분석 방법은 3차원 모델의 구조에 대한 보다 상세한 정보를 제공한다. 특히 의료 분야에 사용될 경우 전문가의 개입을 최소화하여 질병 진단 및 치료 등에 사용될 수 있다. 본 연구에서는, MRI나 CT 영상으로부터 생성된 3차원 매개변수형 모델을 이용하여 유사 모델 집단을 대표하는 통계 형상을 구축한 후, SVM (Support Vector Machine) 학습 알고리즘을 이용하여 두 집단간 형상 차이를 분석한다. 3차원 형상에 대한 신속한 시각적 이해와 직관적 조작감은 물체 표면의 형상 변화를 분석하는데 효과적으로 사용될 수 있다. 본 논문에서는 물체 조작 및 관찰 등의 작업을 수행할 때, 햅틱 피드백과 스테레오 비쥬얼 큐를 혼합한 인터랙션 기법을 사용하여 공간감과 깊이감을 향상시켜 형상 분석 결과를 효과적으로 분석한다. 본 연구에서는 해마, 관상 동맥, 뇌와 같은 인체 장기를 실험 데이터로 사용하여 제안한 SVM 기반의 분석 방법과 인터랙션 환경의 성능을 평가한다. 본 연구에서 구현한 SVM 기반 이진 분류기는 두 집단간 형상 차이를 효과적으로 분석하며, 또한 다중 감각 인터랙션은 사용자가 분석 결과를 관찰하고 카메라 및 형상을 효율적으로 조작하는 데 도움을 준다.