• 한국기계가공학회지
• /
• 제8권4호
• /
• pp.34-40
• /
• 2009

The effect of frictional factors on thermal stress and deformation volume of motorcycle brake disk was studied by using a disk-on-pad type friction tester. It has an effect on the frictional factor such as applied load, sliding speed, sliding distance and number of ventilated disk hole. However, it is difficult to know the mutual relation of these factors. In this study, thermal stress and deformation volume by using design of experiment with 4 elements were investigated for thermal analysis with regression analysis. Thermal stress and thermal deformation are obtained by the application of temperature from mechanical test. From this study, the result showed that the motorcycle brake disk with ventilated hole 3 had the most excellent thermal stress and deformation volume. The regression equation had a trust rate of 95% for the prediction of thermal stress and deformation volume of motorcycle brake disk was composed.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2011년도 추계학술발표대회
• /
• pp.1256-1257
• /
• 2011

온라인 사회연결망의 신뢰 관계를 활용한 양질의 서비스를 개발하기 위해서는 신뢰 관계가 존재하지 않는 사용자들 간의 신뢰 관계를 예측할 수 있는 모델에 대한 연구가 필요하다. 신뢰 관계 예측 모델의 정확도를 향상시키기 위해서는 사용자 인정 기준을 통해 대상 사용자를 선정할 필요가 있다. 본 논문에서는 사용자 인정 기준 변화에 따른 신뢰 관계 예측의 정확도 변화를 분석하여 사용자 인정 기준을 강화하여 신뢰 예측 결과의 정확도를 향상시킬 수 있음을 보인다. 본 연구의 결과가 최적의 사용자 인정 기준을 찾기 위한 실마리를 제공할 수 있을 것으로 기대한다.

• 한국정보통신학회논문지
• /
• 제22권1호
• /
• pp.17-25
• /
• 2018

최근 ICT기술의 발전과 스마트 기기의 급격한 보급으로 엄청난 양의 정보가 생성되고 있다. 추천 시스템은 과도한 정보제공(information overload)으로부터 정보 수용자의 적절한 판단을 도와주고, 정보 제공자에게는 기업의 이윤과 업체홍보 효과를 증대 시킬 수 있는 해결책으로 등장하였다. 추천 시스템은 다양한 접근법으로 구현이 가능하지만, 소셜 네트워크 정보로 성능을 향상시킬 수 있는 방법으로 제시되었다. 그러나 추천 시스템 내의 사용자간에 형성되는 신뢰 클러스터의 정보를 활용하는 방안은 연구되지 못하였다. 본 논문에서는 온라인 리뷰에서 생성되는 클러스터에서 클러스터 내부 객체 간 영향성과 트러스터-트러스티 간 정보를 이용하여 추천 시스템의 성능을 향상시키는 방식을 제안하였다. 제안하는 방식을 구현하고 실제 데이터를 활용하여 실험한 결과 기존의 방식들보다 예측 정확도가 향상됨을 확인하였다.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• 제42권
• /
• pp.37.1-37.13
• /
• 2020

Background: The human mandible is variable in shape, size and position and any deviation from normal can affect the facial appearance and dental occlusion. Objectives: The objectives of this study were to determine whether the Sassouni cephalometric analysis could help predict two-dimensional mandibular shape in humans using cephalometric planes and landmarks. Materials and methods: A retrospective computerised analysis of 100 lateral cephalometric radiographs taken at Kingston Hospital Orthodontic Department was carried out. Results: Results showed that the Euclidean straight-line mean difference between the estimated position of gonion and traced position of gonion was 7.89 mm and the Euclidean straight-line mean difference between the estimated position of pogonion and the traced position of pogonion was 11.15 mm. The length of the anterior cranial base as measured by sella-nasion was positively correlated with the length of the mandibular body gonion-menton, r = 0.381 and regression analysis showed the length of the anterior cranial base sella-nasion could be predictive of the length of the mandibular body gonion-menton by the equation 22.65 + 0.5426x, where x = length of the anterior cranial base (SN). There was a significant association with convex shaped palates and oblique shaped mandibles, p = 0.0004. Conclusions: The method described in this study can be used to help estimate the position of cephalometric points gonion and pogonion and thereby sagittal mandibular length. This method is more accurate in skeletal class I cases and therefore has potential applications in craniofacial anthropology and the 'missing mandible' problem in forensic and archaeological reconstruction.

• 경영정보학연구
• /
• 제22권1호
• /
• pp.1-16
• /
• 2020

블록체인은 시스템에 참여하는 사용자들 사이에 믿음을 확보할 수 있는 기술로 떠오르고 있다. 블록체인에 대한 관심이 높아지면서 선행 연구들은 블록체인 기술에 관련된 암호화폐와 적용방안에 대한 연구를 주로 수행하였다. 반면에 블록체인의 안정적인 구동에 대한 연구는 크게 주목하지 않았다. 대표적으로 블록체인의 엉클블록은 블록체인 시스템의 안정적 구동에 중요한 역할을 담당함에도 불구하고 이에 대한 연구는 거의 수행되지 않았다. 이러한 인식을 기반으로 본 연구는 블록체인 정보와 거시 경제 요인들을 활용하여 블록체인의 엉클블록을 머신러닝 기법으로 예측하고자 하였다. 인공신경망, 서포트벡터머신 분석 결과, 블록체인 정보와 거시 경제 요인들은 블록체인의 엉클블록 예측에 기여하는 것으로 나타났다. 또한, 블록체인 정보만을 활용한 인공신경망은 엉클블록의 발생을 예측하는데 가장 우수한 성능을 제공하는 것으로 나타났다. 본 연구는 정보시스템 분야에서 블록체인 연구를 주도하고 기여할 수 있는 방안을 제시한다.

• The Journal of the Acoustical Society of Korea
• /
• 제27권2E호
• /
• pp.45-50
• /
• 2008

In recent years, quantitative ultrasound (QUS) technologies have played a growing role in the diagnosis of osteoporosis. Most of the commercial bone somometers measure speed of sound (SOS) and/or broadband ultrasonic attenuation (EUA) at peripheral skeletal sites. However, the QUS parameters are purely empirical measures that have not yet been firmly linked to physical parameters such as bone strength or porosity. In the present study, the theoretical models for wave propagation in cancellous bone, such as the Biot model, the stratified model, and the modified Biot-Attenborough (MBA) model, were applied to predict the dependence of phase velocity on porosity in cancellous bone. The optimum values for the input parameters of the three models in cancellous bone were determined by comparing the predictions with the previously published measurements in human cancellous bone in vitro. This modeling effort is relevant to the use of QUS in the diagnosis of osteoporosis because SOS is negatively correlated to the fracture risk of bone, and also advances our understanding of the relationship between phase velocity and porosity in cancellous bone.

• Journal of Microbiology and Biotechnology
• /
• 제14권1호
• /
• pp.74-80
• /
• 2004

Aspergillus fumigatus is one of the most common fungi in the human environment, both in-doors and out-doors. It is the main causative agent of invasive aspergillosis, a life-threatening mycosis among immunocompromised patients. The genome has been sequenced by an international consortium, including the Wellcome Trust Sanger Institute (U.K.) and The Institute for Genomic Research (TIGR, U.S.A.), and a ten times whole genome shotgun sequence assembly has been made publicly available. In this study, we identified tricarboxylic acid (TCA) cycle enzymes of A. fumigatus by comparative analysis with four other fungal species. The open reading frames showed high amino acid sequence similarity with the other fungal citric acid enzymes and well-conserved functional domains. All genes present in Saccharomyces cerevisiae, Schizosaccharomyces pombe, Candida albicans, and Neurospora crassa were also found in A. fumigatus. In addition, we identified four A. fumigatus genes coding for enzymes in the glyoxylate shunt, which may be required for fungal virulence. The architecture of multi-gene encoded enzymes, such as isocitrate dehydrogenase, 2-ketoglutarate, succinyl-CoA synthetase, and succinate dehydrogenase was well conserved in A. fumigatus. Furthermore, our results show that genes of A. fumigatus can be detected reliably using GlimmerM.

• International Journal of Internet, Broadcasting and Communication
• /
• 제16권1호
• /
• pp.203-214
• /
• 2024

This study seeks to provide implications for domestic-related policies through exploratory analysis research to support AI-based policy decision-making. The following should be considered when establishing an AI-based decision-making model in Korea. First, we need to understand the impact that the use of AI will have on policy and the service sector. The positive and negative impacts of AI use need to be better understood, guided by a public value perspective, and take into account the existence of different levels of governance and interests across public policy and service sectors. Second, reliability is essential for implementing innovative AI systems. In most organizations today, comprehensive AI model frameworks to enable and operationalize trust, accountability, and transparency are often insufficient or absent, with limited access to effective guidance, key practices, or government regulations. Third, the AI system is accountable. The OECD AI Principles set out five value-based principles for responsible management of trustworthy AI: inclusive growth, sustainable development and wellbeing, human-centered values and fairness values and fairness, transparency and explainability, robustness, security and safety, and accountability. Based on this, we need to build an AI-based decision-making system in Korea, and efforts should be made to build a system that can support policies by reflecting this. The limiting factor of this study is that it is an exploratory study of existing research data, and we would like to suggest future research plans by collecting opinions from experts in related fields. The expected effect of this study is analytical research on artificial intelligence-based decision-making systems, which will contribute to policy establishment and research in related fields.

• 한국산학기술학회논문지
• /
• 제10권11호
• /
• pp.3465-3472
• /
• 2009

본 병원에 축적된 의무기록과 데이터베이스에 있는 퇴원 환자 정보를 이용하여 이탈에 영향을 미치는 특성을 파악하여 활용 가능한 예측모형을 제시하고자 한다. 외래진료 방문환자 3,503명 중 충성고객 2,118명 60.5%, 이탈 고객 1,385명 39.5%을 추출하여 분석에 사용하였다. 생존한 변수는 성별, 연령(연령대), 지역, 보험구분, 입원경로, 진료과, 퇴원과, 퇴원형태, 협진여부, 수술여부, 진료예약여부, 환자구분을 기반으로 예측모형을 만들었다. 로지스틱 회귀분석을 실시한 결과 66.0%의 정확도를 나타냈고, 신경망을 통하여 예측한 결과 분석용 결과는 정분율은 69.79%이고, 검정용 결과 정분율은 63.64%였다. CHAID를 통하여 예측한 결과 분석용 결과 정분율을 83.75% 이고, 검정용 결과 정분율은 42.74%였다. 예측 모형을 활용한 이탈고객을 위한 관리와 병원의 신뢰를 높여야 할 것이다.

• 한국조리학회지
• /
• 제10권1호
• /
• pp.32-45
• /
• 2004

This study is about choice attribution of customers make in food and beverage events. The researcher provided practical plans to uplift food-related enterprises and activate management through surveys and positive analyses, targeting customers who use food services. First of all, all event plans must include customer demands, social changes, special qualities of the business, and market research. Second, low demand season must be customers will be induced to the events. Third, prediction for market variable and solutions must be thoroughly examined and plans should look into the future to maintain a long period of time. Fourth, sufficient communication between planners and employees should be made before the event starts, so that food and beverage businesses can gain trust and quality of event services.Fifth, immaterial service and visible goods/menus in business of food and beverage events must be closely matched. Sixth, menus introducing a variety of merchandise, quality of nutrition and health of the business should be developed. Also, events from countries(regions) should be hold to create a market of cultural exchange. Seventh, for hereafter event plans, feedbacks are needed concerning customers needs and demands through customer care, after the food and beverage events. Eight, faculty management for convenience, kindness, safety, and life preserver accommodations in parking areas must be made, as automobiles are necessaries for people in Mycar era. The ninth, off-line and on-line care through on-line business construction and production of homepage must be done, due to the fact that even the well-made events are bound to fail if they are not delivered to the customers.