• 반도체디스플레이기술학회지
• /
• 제19권3호
• /
• pp.77-81
• /
• 2020

In this paper, we create scaffold defect classification models using machine learning based data. We extract the characteristic from collected scaffold external images using USB camera. SVM, KNN, MLP algorithm of machine learning was using extracted features. Classification models of three type learned using train dataset. We created scaffold defect classification models using test dataset. We quantified the performance of defect classification models. We have confirmed that the SVM accuracy is 95%. So the best performance model is using SVM.

• 수산해양교육연구
• /
• 제23권1호
• /
• pp.105-115
• /
• 2011

Education is focused on how to nurture creative problem-solving skills talent in rapidly changing information society. The algorithm education of computer science is effective in improvement of students' logical thinking and problem solving capability. However, the algorithm education is very difficult to teach in elementary students level. Because it is difficult to understand abstract characteristic of algorithm. Therefore we developed educational contents based on the principle of the algorithm for improve students' logical thinking and problem-solving capability in this study. And educational contents contain interesting elements of the game. So, students will be interested in algorithm learning and participate actively through developed educational contents. Furthermore, students' creative problem-solving capability may improve through algorithm learning.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2021년도 가을 학술논문 발표대회
• /
• pp.30-31
• /
• 2021

This study aims to build a deep learning model that can predict the value of concrete mixing properties according to a given concrete strength value. A model was created for a total of 1,291 concrete data, including 8 characteristics related to concrete mixing elements and environment, and the compressive strength of concrete. As the deep learning model, DNN-3L-256N, which showed the best performance on the prior study, was used. The average value for each characteristic of the data set was used as the initial input value. In results, in the case of 'curing temperature', which had a narrow range of values in the existing data set, showed the lowest error rate with less than 1% error based on MAE. The highest error rate with an error of 12 to 14% for fly and bfs.

• Nuclear Engineering and Technology
• /
• 제55권10호
• /
• pp.3515-3524
• /
• 2023

A reinforcement learning framework is proposed for the control problem of outlet steam pressure of the once-through steam generator(OTSG) in this paper. The double-layer controller using Proximal Policy Optimization(PPO) algorithm is applied in the control structure of the OTSG. The PPO algorithm can train the neural networks continuously according to the process of interaction with the environment and then the trained controller can realize better control for the OTSG. Meanwhile, reinforcement learning has the characteristic of difficult application in real-world objects, this paper proposes an innovative pretraining method to solve this problem. The difficulty in the application of reinforcement learning lies in training. The optimal strategy of each step is summed up through trial and error, and the training cost is very high. In this paper, the LSTM model is adopted as the training environment for pretraining, which saves training time and improves efficiency. The experimental results show that this method can realize the self-adjustment of control parameters under various working conditions, and the control effect has the advantages of small overshoot, fast stabilization speed, and strong adaptive ability.

• 인터넷정보학회논문지
• /
• 제25권1호
• /
• pp.147-155
• /
• 2024

현재 4차 산업 혁명은 다른 혁명처럼 인류에게 커다란 변화와 새로운 삶을 가져다주고 있으며, 특히 빅데이터, 인공지능, ICT 등 다양한 기술들을 합쳐 응용할 수 있는 드론에 대한 수요와 활용도가 증가하고 있다. 최근에는 러시아-우크라이나 전쟁, 북한의 대남 정찰 등 위험한 군사 작전 및 임무를 수행하는 데 많이 사용되고 있으며 드론에 대한 수요와 활용도가 높아짐에 따라 드론의 안전성과 보안에 대한 우려가 커지고 있다. 현재 드론에 관련된 무선 통신 이상 탐지, 센서 데이터 이상 탐지 등 다양한 연구가 진행되고 있지만, 무선 주파수 특성 데이터를 사용하여 위협을 실시간으로 탐지하는 연구는 미비하다. 따라서, 본 논문에서는 실제 환경과 유사한 HITL(Hardware In The Loop) 시뮬레이션 환경에서 드론이 미션을 수행하는 동안 지상 제어 시스템과 통신하면서 발생하는 무선 주파수 신호 특성 데이터를 수집하여 특성 데이터가 정상 신호 데이터인지 비정상 신호 데이터인지 판단하는 연구를 진행하였다. 또한, 드론이 미션을 수행하는 중 실시간으로 위협 신호를 탐지할 수 있는 비지도 학습 기반의 위협 탐지 시스템 및 최적의 임계값을 제안한다.

• 대한전자공학회논문지SD
• /
• 제38권8호
• /
• pp.1-12
• /
• 2001

본 연구에서는 MFSFET (Metal-Ferroelectric-Semiconductor FET) 소자의 모델링을 바탕으로 adaptive learning 회로를 설계하고, 그 수치적인 결과를 분석하였다. Adaptive learning 회로에서 출력주파수는 MFSFET 소자의 소스-드레인 저항과 캐패시턴스에 반비례하는 특성을 보여주었다. Short pulse 수에 따른 포화드레인 전류곡선은 강유전체의 분극반전 특성과 유사함을 확인할 수 있었고, 이는 강유전체 분극이 MFSFET 소자의 드레인 전류조절에 핵심적인 요소로 작용한다는 사실을 의미한다. 다음으로 MFSFET 소자의 드레인 전류조절에 핵심적인 요소로 작용한다는 사실을 의미한다. 다음으로 MFSFET 소자의 소스-드레인 저항으로부터 dimensionality factor 와 adaptive learning 회로의 펄스 수에 따른 출력주파수 변화를 분석하였다. 이 특성으로부터, adaptive learning 회로의 주파수변조 특성 즉, 입력펄스의 진행에 따라 출력펄스의 점진적인 주파수 변화를 의미하는 adaptive learning 특성을 명화하게 확인할 수 있었고, 뉴럴 네트워크에서 본 회로가 뉴런의 시넵스 부분에 효과적으로 사용될 수 있음을 입증하였다.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2008년도 학술대회 2부
• /
• pp.64-70
• /
• 2008

Computer Assisted Learning (CAL) is quite different from in many aspects. CAL provides individualistic learning environment and facilitates autonomy of the learner. Thus the learners who uses CAL program has more sense of control and engages in more strategic learning than conventional learning environment. In this experiment, we used KORI (KORea university intelligent agent) which is a new type of ITS adopting TA (Teachable Agent) that fosters learning by teaching, So, we investigated the critical motivational factor that have influences in CAL learning and the effects of reward in CAL are another area of our interest. Thus, we divided two conditions that presence of reward and absence of reward. The 174 elementary school students(5th) were participated and they are randomly assigned the one of the reward conditions. Before entering the experimental instruction, all participants measured about metacognition, self-efficacy and goal orientation questionnaire as independent variables. Then, Participants were instructed of method of using KORI program and asked to study for ten days with KORI program at least 20 minutes everyday in their home, about 10 days. After 10 days, they were rated interest and comprehension. Regression results suggest that regardless of the presence of reward, metacognition is a positive predictor in interestingness. It indicate that metacognitive skills are required in CAL learning situation irrespective of reward. But on comprehension in the absence of reward, only self- efficacy appeared to be a positive predictor. In the presence of reward, performance goal orientation showed as a negative predictor of comprehension, whereas self-efficacy was a positive predictor. This result suggest that presence of reward especially interferes learning process of performance goal orientation in CAL learning situation. It could be interpreted that reward interferes the learning process of performance goal orientation by debilitating intrinsic motivation.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제10권5호
• /
• pp.155-162
• /
• 2021

n-스텝 시간차 학습은 몬테카를로 방법과 1-스텝 시간차 학습을 결합한 것으로, 적절한 n을 선택할 경우 몬테카를로 방법과 1-스텝 시간차 학습보다 성능이 좋은 알고리즘으로 알려져 있지만 최적의 n을 선택하는 것에 어려움이 있다. n-스텝 시간차 학습에서 n값 선택의 어려움을 해소하기 위해, 본 논문에서는 Q의 과대평가가 초기 학습의 성능을 높일 수 있다는 특징과 Q ≈ Q^* 경우, 모든 n-스텝 누적 보상이 비슷한 값을 가진다는 성질을 이용하여 1 ≤ k ≤ n에 대한 모든 k-스텝 누적 보상의 최댓값과 평균으로 구성된 새로운 학습 타겟인 Ω-return을 제안한다. 마지막으로 OpenAI Gym의 Atari 게임 환경에서 n-스텝 시간차 학습과의 성능 비교 평가를 진행하여 본 논문에서 제안하는 알고리즘이 n-스텝 시간차 학습 알고리즘보다 성능이 우수하다는 것을 입증한다.

• 한국농촌건축학회논문집
• /
• 제13권1호
• /
• pp.21-28
• /
• 2011

The main purpose of this study is to analyze construction of space through the investigation of the cases of school community library through Gangneung-si's life-learning project and the findings from the analysis could be summarized as below. Firstly, most space used for school community library has the size of two classes in school on average and locals use generally space for reference and learning at school community library. Secondly, the construction of space of school community library is categorized into one for book-returning, references, reading, group learning and information, and an audio-visual space is also used for group learning and reading. A space for book-returning has features based on the location of its entrance and a space for reading features stand-up and sitting-on space considering size and usability. And a space for group learning has the feature of space planning that makes it possible for local people to get library programs and seminars and a space for information shows its feature of space planning that uses the wall.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권6호
• /
• pp.2527-2545
• /
• 2016

Measuring the similarity of given samples is a key problem of recognition, clustering, retrieval and related applications. A number of works, e.g. kernel method and metric learning, have been contributed to this problem. The challenge of similarity learning is to find a similarity robust to intra-class variance and simultaneously selective to inter-class characteristic. We observed that, the similarity measure can be improved if the data distribution and hidden semantic information are exploited in a more sophisticated way. In this paper, we propose a similarity learning approach for retrieval and recognition. The approach, termed as LDA-FEK, derives free energy kernel (FEK) from Latent Dirichlet Allocation (LDA). First, it trains LDA and constructs kernel using the parameters and variables of the trained model. Then, the unknown kernel parameters are learned by a discriminative learning approach. The main contributions of the proposed method are twofold: (1) the method is computationally efficient and scalable since the parameters in kernel are determined in a staged way; (2) the method exploits data distribution and semantic level hidden information by means of LDA. To evaluate the performance of LDA-FEK, we apply it for image retrieval over two data sets and for text categorization on four popular data sets. The results show the competitive performance of our method.