• Proceedings of the Korean Nuclear Society Conference
• /
• 1997.10a
• /
• pp.38-43
• /
• 1997

이 논문의 목적은 신경망 이론에 의한 노심특성평가 모델과 유전자 기법을 써서 가압경수로 노심의 최적화 재장전 모형을 결정하는데 있다. 이를 위해 OLL(Optimization Layer by Layer)신경망을 구축하고 이를 영광 3호기 재장전주기 노심특성(특히 연료집합체 출력분포와 임계붕산농도)을 예측할 수 있도록 훈련하여 영광3호기 재장전주기 특성 해석용 OLL 신경망을 만들었다. 그리고 통상의 유전자 기법을 활용하여 매세대당 150개의 장전모형들을 생산하고 이들을 대상으로 1000세대에 걸친 유전자 기법에 의한 최적화 과정을 통해 영광 3호기 노심의 평형주기 최적 모형을 결정하였다.

• Journal of the Korea Convergence Society
• /
• v.9 no.5
• /
• pp.21-26
• /
• 2018

In this study, HRD-NET (http://hrd.go.kr), a vocational and training integrated computer network operated by the Ministry of Employment and Labor, is used to confirm whether job training information required by job seekers is being provided smoothly The question bulletin board was extracted using 'R' program which is optimized for big data technique. Therefore, the effectiveness, appropriateness, visualization, frequency analysis and association analysis of the vocational training system were conducted through this, The results of the study are as follows. First, the issue of vocational training card, video viewing, certificate issue, registration error, Second, management and processing procedures of learning cards for tomorrow 's learning cards are complicated and difficult. In addition, it was analyzed that the training cost system and the refund structure differentiated according to the training occupation, the process, and the training institution in the course of the training. Based on this paper, we will study not only the training system of the Ministry of Employment and Labor but also the improvement of the various training computer system of the government department through the analysis of big data.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.12
• /
• pp.459-465
• /
• 2017

This research aims to assess whether neurofeedback training can reduce theta waves in adolescents. The experiment was conducted on 35 early youths living in Gyeonggi-do at youth counseling centers during April-October. According to circumstances and opinions of participants in the pre-brain analysis, they were classified into a non-training group (A), 12-week training group (B), and 24-week training group (C), containing 10, 15, and 10 members, respectively. EEG measurement and neurofeedback training was performed using the prefrontal 2-channel NeuroharmonyS and Brain Optimization program. EEG data was processed utilizing Brain Analysis ver1.3. Deducted data was converted to SPSS 21.0 to enable statistical processing. As a strategy to reduce theta through the Beta increase training, we applied the appropriate Alpha, SMR, Beta low reward training to the individual. Study results confirmed that theta waves of adolescents decreased through the prefrontal neurofeedback training. Groups (B) and (C) exhibited a greater decrease in theta waves compared with the control group.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.399-401
• /
• 2022

A definitive diagnosis to classify the breast malignancy status may be achieved by microscopic analysis using surgical open biopsy. However, this procedure requires experts in the specializing of histopathological image analysis directing to time-consuming and high cost. To overcome these issues, deep learning is considered practically efficient to categorize breast cancer into benign and malignant from histopathological images in order to assist pathologists. This study presents a pre-trained convolutional neural network model architecture with a 100% fine-tuning scheme and Adagrad optimizer to classify the breast cancer histopathological images into benign and malignant using a 40× magnification BreaKHis dataset. The pre-trained architecture was constructed using the InceptionResNetV2 model to generate a modified InceptionResNetV2 by substituting the last layer with dense and dropout layers. The results by demonstrating training loss of 0.25%, training accuracy of 99.96%, validation loss of 3.10%, validation accuracy of 99.41%, test loss of 8.46%, and test accuracy of 98.75% indicated that the modified InceptionResNetV2 model is reliable to predict the breast malignancy type from histopathological images. Future works are necessary to focus on k-fold cross-validation, optimizer, model, hyperparameter optimization, and classification on 100×, 200×, and 400× magnification.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.10
• /
• pp.483-488
• /
• 2016

In this paper, we propose a prototype-based classification learning by using the nearest-neighbor rule. The nearest-neighbor is applied to segment the class area of all the training data with hyperspheres, and a hypersphere must cover the data from the same class. The radius of a hypersphere is computed by the mid point of the two distances to the farthest same class point and the nearest other class point. And we transform the prototype selection problem into a set covering problem in order to determine the smallest set of prototypes that cover all the training data. The proposed prototype selection method is designed by a greedy algorithm and applicable to process a large-scale training set in parallel. The prediction rule is the nearest-neighbor rule and the new training data is the set of prototypes. In experiments, the generalization performance of the proposed method is superior to existing methods.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.291-291
• /
• 2021

모형이 갖는 불확실성의 정량화나 매개변수의 최적화는 계산시간의 기하급수적인 증가를 가져온다. 계산시간의 효율성을 극대화할 수 있는 기법으로 최근 대체모형이 개발되었으며, 다양한 분야에서 적용되고 있다. 그러나 대체모형은 훈련된 데이터 공간에서 크게 벗어난 극한 사상를 정확하게 모의하기는 어려운 단점이 있다. 본 연구는 이와 같은 대체모형의 단점을 개선할 수 있는 새로운 PCK(polynomial chaos-krigging) 기법을 제시한다. PCK는 PCE(polynomial chaos expansion) 기법과 OK(ordinary krigging) 기법을 결합한 것이며, PCK의 효과는 기존의 PCE 및 OK 모형의 결과와 비교하여 입증하였다. 본 연구의 분석 결과는 다음과 같다. (1) PCK는 더 적은 수의 훈련 샘플만으로도 원래 모형을 더 정확하게 대체할 수 있다. (2) 원래 훈련 샘플보다 약 3배 더 큰 극한사상을 모의했을 때, PCE와 OK는 예측이 실패하였지만, PCK의 예측은 정확하였다. (3) 민감도 분석 결과 PCK의 매개변수 특성과 거동이 PCE 및 OK보다 원래 모형의 특성과 거동에 더 일치한다. 본 연구에서는 3개의 대체모형의 결과를 원래모형의 결과와 비교하였으며 그 적용성을 극한강우에 대해 검토하였다. 일반적으로 훈련 샘플의 범위와 비슷한 강우사상에 대해서는 모든 대체모형의 결과가 우수하였으나, 훈련 샘플의 범위에서 벗어난 극한 사상의 모의는 PCK만 적용이 가능하였다. 제안된 대체모형은 극한사상의 예측에 있어 기존 대체모형보다 매우 향상된 정확도를 제공함을 확인할 수 있었다.

• Journal of Internet Computing and Services
• /
• v.25 no.4
• /
• pp.23-37
• /
• 2024

Digital twin is an M&S (Modeling and Simulation) technology designed to solve or optimize problems in the real world by replicating physical objects in the real world as virtual objects in the digital world and predicting phenomena that may occur in the future through simulation. Digital twins have been elaborately designed and utilized based on data collected to achieve specific purposes in large-scale environments such as cities and industrial facilities. In order to apply this digital twin technology to real life and expand it into user-customized service technology, practical but sensitive issues such as personal information protection and personalization of simulations must be resolved. To solve this problem, this paper proposes a federated learning-based accelerated client training method (FACTS) for personalized digital twins. The basic approach is to use a cluster-driven federated learning training procedure to protect personal information while simultaneously selecting a training model similar to the user and training it adaptively. As a result of experiments under various statistically heterogeneous conditions, FACTS was found to be superior to the existing FL method in terms of training speed and resource efficiency.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.283-285
• /
• 2023

이상 탐지는 데이터에서 일반적인 범주에서 크게 벗어나는 인스턴스 또는 패턴을 식별하는 중요한 작업이다. 본 연구에서는 시계열 데이터의 특징 추출을 위한 비지도 학습 기반 방법과 정규화 흐름의 결합을 통한 이상 탐지 프레임워크를 제안한다. 특징 추출기는 1차원 합성곱 신경망 기반의 오토인코더로 구성되며, 정상적인 시퀀스로만 구성된 훈련 데이터를 압축하고 복원하는 과정을 통해 최적화된다. 추출된 시계열 데이터의 특징 맵은 가능도를 최대화하도록 훈련된 정규화 흐름의 입력으로 사용된다. 이와 같은 방식으로 훈련된 이상 탐지 시스템은 테스트 샘플에 대한 이상치를 계산하며, 최종적으로 임계값과의 비교를 통해 이상 여부를 예측한다. 성능 평가를 위해 시계열 이상 탐지를 위한 공개 데이터셋을 이용하여 공정하게 이상 탐지 성능을 비교하였으며, 실험 결과는 제안하는 정규화 흐름 기법이 시계열 이상 탐지 시스템에 활용될수 있는 잠재성을 시사한다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.8
• /
• pp.1765-1772
• /
• 2015

In the underwater acoustic communication channels, multipath reflection become the cause of obstacle. Generally, equalizer has been applied to overcome these problems. In this paper, the method was proposed to optimize tap-length of decision feedback equalizer using genetic algorithm. After inputting feed-forward filter length and feed-back filter length as genetic information of the genetic algorithm, it optimize tap-length using BER(bit error rate) calculation in accordance with object function. The object function consist of decision feedback equalizer and BER calculation. For the purpose of BER calculation in the object function, the method was proposed to optimize the tap-length of decision feedback equalizer with genetic algorithm using preamble signals. As a result of experiments, the optimized BER is 0.0355 for signals which were received through a 25m receiver and which were applied to calculate BER merely using preamble signals in object function. When all data were used to calculate BER in object function, the optimized BER is 0.0215.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.25 no.6
• /
• pp.726-734
• /
• 2019

The current maritime industry is expected to have a significant impact on the role of maritime-related technologies and systems, especially seafarers, in the rapidly changing Fourth Industrial Revolution. The Maritime Autonomous Surface Ship (MASS) aims to reduce the number of safety accidents and improve seafarers' working environment. With regard to MASS, the International Maritime Organization has been trying to minimize unexpected impact in the maritime education and training sector by establishing international conventions such as the Standards of Training, Certification and Watchkeeping for Seafarers. However, domestic designated educational institutions have not yet established an education and training scheme to develop seafarers who will be on board for MASS. Therefore, this paper reviews the technology of MASS, analyzes the changes in education and training in order to upgrade the qualifications, and suggests the competencies of smart seafarers equipped with the integrated management ability required for Artificial Intelligence, Big Data, Cybersecurity, and the Digital System Revolution through education and training. In addition, this study provides basic information for the education and training of seafarers who are optimized for the rapidly changing technological environment.