• Proceedings of the KIEE Conference
• /
• 2002.07d
• /
• pp.2846-2849
• /
• 2002

본 논문은 뉴로-퍼지 시스템에서의 규칙 선택 및 모델 학술에 대하여 EM 알고리즘을 기반으로 하는 구조 동정을 제안한다. 뉴로-퍼지 모델링에서의 초기 파라미터가 학습과정에서의 모델 성능에 큰 영향을 주고 있다. 주어진 데이터에 근거한 파라미터 추정에는 다양한 방법들이 소개되고 응용되어져 왔는데 이전 연구들에서 볼 수 있는 HCM, FCM 등은 데이터와의 유클리디언 거리를 최소화하는 중심점을 파라미터로 선택하는 등의 방법과 퍼지 균등화 등은 데이터의 확률 밀도함수를 이용하여 파라미터를 추정하였다. 제안된 방법에서는 데이터에서의 Maximum Likelihood Estimator를 기반으로 하는 방법으로 EM 알고리즘을 이용하였다. 초기 파라미터의 결정에서 EM 알고리즘을 이용하여 뉴로-퍼지 모델의 전제부 소속함수 파라미터 추정을 실시한다. EM 알고리즘을 이용한 퍼지 모델의 특징으로는 전제부가 클러스터링에 의하여 생성되므로 입력의 차원이나 소속함수의 수가 증가하여도 규칙의 수는 증가하지 않는다. 이를 자동차 MPG 예제를 통하여 제안된 방법의 유용성을 보이고자 한다.

• Journal of Korean Society of Rural Planning
• /
• v.11 no.1 s.26
• /
• pp.35-43
• /
• 2005

Practically, it is very hard to investigate the rural villages in North Korea. The Korean rural village in Yanbian is proximate from North Korea and most of inhabitants of the village have been emigrated from Hamgyeong-Do and Pyeongan-Do. In this sense, it can be meaningful to study Korean villages in Yanbian is an alternative approach toward researches of North Korean rural villages. In this study, spatial characteristics of the Ryongsan village, Ryongjeong city was studied as a representative korean rural village in Yanbian. The Ryongsan village has been formed based on the traditional culture and custom of North Korea. The case of planning and design is provided through drawings of components for planning of rural villages based on the results of spatial characteristics analysis.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.598-601
• /
• 2019

최근 딥러닝과 신경망 기술을 바탕으로 한 다양한 번역 시스템이 상용화되고 있다. 그러나 문맥을 고려해야하는 경우나 의료, 법학 등 전문적인 지식을 요구하는 경우, 기존의 번역 시스템으로는 원문 작성자가 의도한 대로 자연스러운 번역 결과물을 얻기 어려운 문제점이 있다. 이로 인해 전문적인 지식을 요구하는 문서는 전문 번역 기관에 의뢰하여 번역이 이루어지고 있으며 소비자의 입장에서는 업체로부터 받은 번역물을 평가 과정 없이 사용할 수밖에 없는 상황이다. 따라서 본 논문에서는 사람이 직접 번역한 결과물을 사용자가 가진 번역 신뢰 점수를 반영하여 평가함으로써 양질의 결과물을 도출하고 번역 활동 기여도에 따라 암호화폐로 보상금을 지급하는 블록체인 기반 보상형 공동 번역 모델을 제안한다. 다수의 번역 평가자들로부터 높은 점수를 받은 우수한 번역 결과물을 도출하고, 번역글 평가에 사용되는 사용자별 신뢰도 점수를 블록체인에 저장하여 조작을 방지함으로써 믿을 수 있는 번역 생태계를 조성할 것으로 기대한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.393-396
• /
• 2008

센서 네트워크에서 병합 질의를 효율적으로 처리하기 위한 다양한 인-네트워크 질의 처리 기법이 제안되었다. 스카이라인 질의는 일반적인 병합 질의와 달리 다차원 데이터에 대한 비교를 요구하므로 인-네트워크 처리가 쉽지 않다. 스카이라인 질의를 에너지 효율적으로 처리하기 위해서 불필요한 데이터의 전송을 제거하는 것이 중요하다. 기존에 제안된 스카이라인 처리 기법은 전체 네트워크에 필터를 배포함으로써 불필요한 데이터 전송을 차단한다. 하지만 많은 False Positive 발생에 따른 불필요한 데이터 전송과 필터 배포시 발생하는 에너지 소모로 인해 네트워크의 수명이 단축된다. 본 논문에서는 필터 배포에 따른 에너지 소모를 줄이기 위한 방법으로 상향식 필터 설정을 통한 스카이라인 질의 처리 기법과 필터링 성능을 향상시키는 기법을 제안한다. 제안하는 기법은 데이터를 수집하는 과정에서 스카이라인 필터테이블(SFT)설정하는 상향식 필터링을 수행한다. 그리고 선-필터링(Pre-filtering) 기법을 통해 필터효과를 증가시킨다. 제안하는 알고리즘의 우수성을 보이기 위해 시뮬레이션을 통해 기존에 제안된 MFTAC기법과 비교하였으며, 그 결과 평균 False Positive가 평균 84.44% 감소하였고, 네트워크 수명이 약 75.99% 증가하였다.

• Journal of the Korean Society of Tobacco Science
• /
• v.23 no.2
• /
• pp.168-178
• /
• 2001

Essential oil in tobacco leaves influences the taste and aroma of cigarette smoke and is important to tobacco quality. This study was conducted to investigate the change in the level of essential oil components during flue-curing process of two flue-cured tobaccos, NC82 and KEl14. Flue-curing process was divided by six steps; harvest stage, the end of yellowing stage, the middle of color fixing stage, the end of color fixing stage, the middle of midrib drying stage, full-cured stage. NC82 in each stage contained 0.28%, 0.30%, 0.35%, 0.36%, 0.40% and 0.42% essential oil, respectively, and KF114 were 0.29%, 0.31%, 0.34%, 0.36%, 0.39% and 0.41%, respectively. Almost all hydrocarbons on the basis of relative peak area were gradually increased in two varieties with curing, neophytadiene content in them was highest at the full-cured stage. Most of alcohols and esters with curing showed a declining trend, but benzyl alcohol was increased in two tobaccos. Ketones were largely increased at the midrib drying stage during the curing process, especially, the most largely increasing constituent was $\beta$-damascenone among them. The content of 2-butylterahydrofuran, heterocyclic compounds, was largely increased at tile color fixing stage. There was no considerable difference between NC82 and KFl14 at the GC profile of essential oil and the pattern of each components during flue-curing process.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.2
• /
• pp.204-211
• /
• 2020

In this paper, we propose an optimal deep learning network architecture for main object location detection through weak supervised learning. The proposed network adds convolution blocks for improving the localization accuracy of the main object through weakly-supervised learning. The additional deep learning network consists of five additional blocks that add a composite product layer based on VGG-16. And the proposed network was trained by the method of weakly-supervised learning that does not require real location information for objects. In addition, Grad-CAM to compensate for the weakness of GAP in CAM, which is one of weak supervised learning methods, was used. The proposed network was tested through the CUB-200-2011 data set, we could obtain 50.13% in top-1 localization error. Also, the proposed network shows higher accuracy in detecting the main object than the existing method.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.12
• /
• pp.1919-1925
• /
• 2022

In this paper, we propose an image-to-image translation approach based on the conditional generative adversarial network for semantic segmentation. Semantic segmentation is the task of clustering parts of an image together which belong to the same object class. Unlike the traditional pixel-wise classification approach, the proposed method parses an input RGB image to its corresponding semantic segmentation mask using a pixel regression approach. The proposed method is based on the Pix2Pix image synthesis method. We employ residual connections-based convolutional neural network architectures for both the generator and discriminator architectures, as the residual connections speed up the training process and generate more accurate results. The proposed method has been trained and tested on the NYU-depthV2 dataset and could achieve a good mIOU value (49.5%). We also compare the proposed approach to the current methods in semantic segmentation showing that the proposed method outperforms most of those methods.

• Journal of applied mathematics & informatics
• /
• v.12 no.1_2
• /
• pp.307-312
• /
• 2003

In this paper, a special incidence structure is designed based on (v, k+1, 1)-configuration, a subset of symmetric balanced incomplete block design, where ν = k$^2$+ k + 1. The time complexity required for desinging this incidence structure is O(equation omitted).

• Journal of the Korean Home Economics Association
• /
• v.41 no.5
• /
• pp.15-39
• /
• 2003

The purpose of this study was to suggest the elderly consumer's consumer informationization education connected with the informationization education for the community by Chung-buk province. Three hundred seventy nine elders were surveyed, with questionnaires from August 12 to September 30, 2002, and a hypothetical model of the consumer informationization education on the basis of the informationization education for the suggestion of the elderly consumer's consumer informationization education was tested. The results were as follows: the perception and the attitude on the informationization education of the elderly were, with a statistical significance, positively related to the perception and the attitude on the consumer informationization education of the elderly. The consumer informationization education (basis, application, attendance, and mission) on the basis of the informationization education (basis and application) in steps was supported as a suggestion for the elderly consumer's consumer informationization education.

• The Korean Journal of Ecology
• /
• v.6 no.4
• /
• pp.250-256
• /
• 1983

Dehiscence of ginseng seeds under the various conditions, growing experiment in the different soil conditions, and growing experiment with extracts obtained by water passed the pot of 4 year ginseng plants were performed. The results obtained are as follows: The survival rate of ginseng plants in the soil of ginseng field sterilized by MIYAZAWA's method was higher than that of the non-sterilized, but lower than that of the control. The survival rate of ginseng plants on the sterilized soil of ginseng field was lower than that on the surrounding soil. The growth of ginseng plants cultivated with the water passed through the pot-planted 4-year-old ginseng plants was strongly inhibited in comparison with the control. Besides biological cause, chemical products excreted from the root of ginseng plants might be the important factors of soil sickness in the successive cultivation of the same plants.