• 한국가스학회지
• /
• 제16권3호
• /
• pp.22-28
• /
• 2012

BPA 공장의 메탄올 분리공정에서 HAZOP 평가를 실시하고, 사고 시나리오로부터 화재 및 폭발 사고의 피해범위를 예측하였다. 그 결과, 화재사고의 피해범위는 50 mm 직경의 안전밸브 토출배관 파열에 의한 제트화재에서는 20 m이었고, 설비가 전파되어 플래쉬화재가 발생되는 경우에는 267 m이었다. 또한 개방공간 증기운 폭발사고의 피해범위는 토출배관 파열에서는 22 m이었고, 설비 전파인 경우에는 542 m이었다. 그리고 최악의 누출 시나리오에 대한 안전대책으로는 메탄올 분리컬럼 내부의 이상압력 상승을 감지할 수 있는 압력계를 2 out of 3 voting으로 설비 상부에 설치하여 주공급라인 상에 설치된 컨트롤밸브와 긴급차단밸브를 동시에 차단할 수 있도록 하여야 한다.

• 한국정밀공학회지
• /
• 제22권9호
• /
• pp.77-84
• /
• 2005

In most vision applications, we are frequently confronted with determining the position of object continuously. Generally, intertwined processes ire needed for target tracking, composed with tracking and control process. Each of these processes can be studied independently. In case of actual implementation we must consider the interaction between them to achieve robust performance. In this paper, the robust real time visual tracking in complex background is considered. A common approach to increase robustness of a tracking system is to use known geometric models (CAD model etc.) or to attach the marker. In case an object has arbitrary shape or it is difficult to attach the marker to object, we present a method to track the target easily as we set up the color and shape for a part of object previously. Robust detection can be achieved by integrating voting-based visual cues. Kalman filter is used to estimate the motion of moving object in 3D space, and this algorithm is tested in a pan/tilt robot system. Experimental results show that fusion of cues and motion estimation in a tracking system has a robust performance.

• 정보처리학회논문지B
• /
• 제11B권4호
• /
• pp.465-470
• /
• 2004

기존의 웹 정보 예측 시스템은 예측에 필요한 정보를 얻기 위하여 사용자 프로파일과 사용자로부터의 명시적 피드백 정보를 필요로 하는 단점이 존재한다. 본 논문에서는 이러한 단점을 극복하고자 웹 사이트에 접속한 고객의 행동을 나타내는 클릭 스트림 데이터와 이를 기반으로 한 사용자의 암시적 피드백 정보를 이용하여 각 사용자가 가장 필요로 하는 웹 정보를 예측한다. 이를 이용하여 관련 정보를 제공할 수 있는 앙상블 SVM을 이용한 동적 웹 정보 예측 시스템을 설계하고 구현하며, 기존의 웹 정보 예측 시스템과 성능 비교를 수행한 결과, 제안된 방법의 우수함이 입증되었다.

• 한국통신학회논문지
• /
• 제17권3호
• /
• pp.247-259
• /
• 1992

본 연구에서는 이동 통신의 AMPS 방식에서 통화를 하기 위해 셀 사이트와 이동체간에 오고 가는 데이터를 분석한 후, 광대역 데이터를 처리하기 위한 시스템을 설계하고 제작하였다. AMPS 방식에서 이동체가 셀 사이트와 통화를 하기 이전에 채널의 상태를 판단하는데 필요한 BUSY / IDLE 비트를 추\ulcorner라흔 회로, 동기 감지 회로, 인터럽트 방식의 데이터 송,수신 회로 및 적은 버퍼 용량을 차지하면서 실시간 처리를 할 수 있는 majority vote, BCH 인코딩 및 디코딩을 하는데 있어서 기존 방법에 따른 계산상의 복잡성을 해결하는 실시간 처리 소프트 웨어를 제안하였다.

• 보건교육건강증진학회지
• /
• 제33권5호
• /
• pp.107-119
• /
• 2016

Objectives: This study aims at examining the association of sense of community with community participation for health promotion in urban areas of Korea. Methods: We analyzed data from 'Community Capacity for Healthy Gangdong Communites' Survey' in 2007. The survey was based on self-reported questionnaires, which were distributed to 1,800 community residents over the age of nineteen in five administrative communities of Gangdong-gu, Seoul, in Korea by using proportionate probability sampling method. We measured 'Sense of community' with four indicators including 'Good neighborhoods', 'Perceived possibility of cooperation', 'Pride of community' and 'Possibility of development' by using 5-point Likert scales. Community participation was measured with the experience rate or the extent of participation by 5-point Likert scales in seven community actions or activities including voting, community program planning, social actions, etc. We examined the association of sense of community with community participation by using regression analyses. Results: This study has shown that sense of community was associated with and made positive impacts on community participation in diverse community actions or activities in urban communities. Conclusions: For promoting community health in urban areas, we can increase community participation more effectively with the efforts of enhancing sense of community.

• 정보처리학회논문지B
• /
• 제10B권6호
• /
• pp.579-586
• /
• 2003

감정요소를 사용한 정보검색시스템은 감정에 기반한 정보검색을 수행하기 위하여 감정시소러스를 구성하였으며 이를 사용한 감정요소추출기를 구현하였다. 감정요소추출기는 기본 5가지 감정 요소를 해당 문서에서 추출하여 문서를 벡터화시킨다. 벡터화시킨 문서들은 k-nearest neighbor, 단순 베이지안 및 상관계수기법을 사용한 2단계 투표방식을 통해 학습하고 분류하였다. 실험결과 분류 방식과 K-means를 이용한 클러스터링에서 감정요소에 기반한 방식이 더 우수하다는 결과와 5,000 단어 미만의 문서 검색에 감정기반 검색이 유리하다는 것을 보였다.

• Journal of Information Processing Systems
• /
• 제10권3호
• /
• pp.443-458
• /
• 2014

An extreme learning machine (ELM) is a recently proposed learning algorithm for a single-layer feed forward neural network. In this paper we studied the ensemble of ELM by using a bagging algorithm for facial expression recognition (FER). Facial expression analysis is widely used in the behavior interpretation of emotions, for cognitive science, and social interactions. This paper presents a method for FER based on the histogram of orientation gradient (HOG) features using an ELM ensemble. First, the HOG features were extracted from the face image by dividing it into a number of small cells. A bagging algorithm was then used to construct many different bags of training data and each of them was trained by using separate ELMs. To recognize the expression of the input face image, HOG features were fed to each trained ELM and the results were combined by using a majority voting scheme. The ELM ensemble using bagging improves the generalized capability of the network significantly. The two available datasets (JAFFE and CK+) of facial expressions were used to evaluate the performance of the proposed classification system. Even the performance of individual ELM was smaller and the ELM ensemble using a bagging algorithm improved the recognition performance significantly.

• Journal of Information Processing Systems
• /
• 제8권1호
• /
• pp.175-190
• /
• 2012

Shuffling is an effective method to build a publicly verifiable mix network to implement verifiable anonymous channels that can be used for important cryptographic applications like electronic voting and electronic cash. One shuffling scheme by Groth is claimed to be secure and efficient. However, its soundness has not been formally proven. An attack against the soundness of this shuffling scheme is presented in this paper. Such an attack compromises the soundness of the mix network based on it. Two new shuffling protocols are designed on the basis of Groth's shuffling and batch verification techniques. The first new protocol is not completely sound, but is formally analyzed in regards to soundness, so it can be applied to build a mix network with formally proven soundness. The second new protocol is completely sound, so is more convenient to apply. Formal analysis in this paper guarantees that both new shuffling protocols can be employed to build mix networks with formally provable soundness. Both protocols prevent the attack against soundness in Groth's scheme. Both new shuffling protocols are very efficient as batch-verification-based efficiency-improving mechanisms have been adopted. The second protocol is even simpler and more elegant than the first one as it is based on a novel batch cryptographic technique.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2005년도 한국컴퓨터종합학술대회 논문집 Vol.32 No.1 (B)
• /
• pp.226-228
• /
• 2005

Predicting the cellular location of an unknown protein gives valuable information for inferring the possible function of the protein. For more accurate Prediction system, we need a good feature extraction method that transforms the raw sequence data into the numerical feature vector, minimizing information loss. In this paper we propose new methods of extracting underlying features only from the sequence data by computing pairwise sequence alignment scores. In addition, we use composition based features to improve prediction accuracy. To construct an SVM ensemble from separately trained SVM classifiers, we propose specificity based weighted majority voting . The overall prediction accuracy evaluated by the 5-fold cross-validation reached $88.53\%$ for the eukaryotic animal data set. By comparing the prediction accuracy of various feature extraction methods, we could get the biological insight on the location of targeting information. Our numerical experiments confirm that our new feature extraction methods are very useful forpredicting subcellular localization of proteins.

• Geomechanics and Engineering
• /
• 제37권6호
• /
• pp.539-554
• /
• 2024

This study explores development of prediction model for seismic site classification through the integration of machine learning techniques with horizontal-to-vertical spectral ratio (HVSR) methodologies. To improve model accuracy, the research employs outlier detection methods and, synthetic minority over-sampling technique (SMOTE) for data balance, and evaluates using seven machine learning models using seismic data from KiK-net. Notably, light gradient boosting method (LGBM), gradient boosting, and decision tree models exhibit improved performance when coupled with SMOTE, while Multiple linear regression (MLR) and Support vector machine (SVM) models show reduced efficacy. Outlier detection techniques significantly enhance accuracy, particularly for LGBM, gradient boosting, and voting boosting. The ensemble of LGBM with the isolation forest and SMOTE achieves the highest accuracy of 0.91, with LGBM and local outlier factor yielding the highest F1-score of 0.79. Consistently outperforming other models, LGBM proves most efficient for seismic site classification when supported by appropriate preprocessing procedures. These findings show the significance of outlier detection and data balancing for precise seismic soil classification prediction, offering insights and highlighting the potential of machine learning in optimizing site classification accuracy.