http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Adaptive neural network ensemble using prediction frequency
Lee Ungki,강남우 한국CDE학회 2023 Journal of computational design and engineering Vol.10 No.4
Neural network (NN) ensembles can reduce large prediction variance of NN and improve prediction accuracy. For highly non-linear problems with insufficient data set, the prediction accuracy of NN models becomes unstable, resulting in a decrease in the accuracy of ensembles. Therefore, this study proposes a prediction frequency-based ensemble that identifies core prediction values, which are core prediction members to be used in the ensemble and are expected to be concentrated near the true response. The prediction frequency-based ensemble classifies core prediction values supported by multiple NN models by conducting statistical analysis with a frequency distribution, which is a collection of prediction values obtained from various NN models for a given prediction point. The prediction frequency-based ensemble searches for a range of prediction values that contains prediction values above a certain frequency, and thus the predictive performance can be improved by excluding prediction values with low accuracy and coping with the uncertainty of the most frequent value. An adaptive sampling strategy that sequentially adds samples based on the core prediction variance calculated as the variance of the core prediction values is proposed to improve the predictive performance of the prediction frequency-based ensemble efficiently. Results of various case studies show that the prediction accuracy of the prediction frequency-based ensemble is higher than that of Kriging and other existing ensemble methods. In addition, the proposed adaptive sampling strategy effectively improves the predictive performance of the prediction frequency-based ensemble compared with the previously developed space-filling and prediction variance-based strategies.
Reliability-Based Design for Market Systems (RBDMS) : Case Study on Electric Vehicle Design
Ungki Lee(이웅기),Namwoo Kang(강남우),Ikjin Lee(이익진) 대한기계학회 2017 대한기계학회 춘추학술대회 Vol.2017 No.11
When designing a product, both engineering uncertainty and market heterogeneity should be considered to reduce the risk of failure in the market. Reliability-based design optimization (RBDO) approach allows decision makers to achieve target confidence in product performance under engineering uncertainty. On the other hand, design for market systems (DMS) approach helps decision makers to find profit-maximized product design under market heterogeneity. This paper proposes a reliability-based design for market systems (RBDMS) framework for electric vehicle (EV) design by integrating RBDO and DMS approaches. In RBDO, the product quality is controlled based on arbitrarily defined reliability. However, from the viewpoint of the company, it is necessary to set the appropriate reliability since the increased reliability increases both consumer satisfaction and manufacturing cost. Therefore, by modeling how reliability affects the customer choice, the optimum reliability which maximizes the profit can be derived. The “reliability” in the framework is used as follows: (1) a decision variable; (2) an attribute that directly influences the customer preference; (3) standards to determine advertised performance; and (4) reliability constraints in engineering model. We optimized and compared four scenarios depending on whether engineering systems are deterministic or probabilistic, and whether a market is homogeneous or heterogeneous. The results show that designing an entry EV model with low reliability is recommended under engineering uncertainty and market heterogeneity, while designing a premium EV model with high reliability is recommended to ensure the robustness of the profit.
‘균형 잡힌’ 방역이라야 지속가능하다 : (포스트-)코로나 위기의 대응과 한국 보건의료 개혁의 과제
정웅기(Ungki Jung),김상준(SangJune Kim),이희영(Heeyoung Lee),최세진(Sejin Choi),김태영(Taeyoung Kim) 한국보건사회학회 2021 보건과 사회과학 Vol.- No.56
본 연구는 코로나19 위기가 장기화되면서 한국이 지속가능한 대응을 하기 위해서는 균형 잡힌 방역으로의 전환이 긴요하다고 주장한다. 우리는 균형의 의미를 두 가지로 이해한다. 첫째, 장기적 관점에서 방역은 기존의 지속적 억제를 고수해서도, 급격한 완화를 추구해서도 지속가능하지 않다. 따라서 중환자 돌봄 역량의 확충을 의료체계 강화의 핵심 목표로 삼는 동시에, 방역 완화에 대비해 무엇보다 학생의 등교권 보장 대책과 요양시설 거주 노인들의 보호 대책을 적극적으로 검토해야 한다. 여기서 선 역량 확보ㆍ후 방역 완화라는 방역 전략이 따라 나온다. 둘째, 방역이 지속가능하려면 미국과 유럽처럼 개인의 자유와 권리에 대한 침해에 극도로 소극적이어서도, 한국처럼 그에 관한 유의미한 사회적 토론이나 숙의를 경유하지 않은 채 일방향적으로 진행되어서도 안 된다. 특히 이 글에서는 대중의 인지적 특성과 감정의 역할을 고려한 보다 효과적인 리스크 커뮤니케이션 대책과 방역 정책에 대한 대중적 관여를 제도적으로 보장할 수 있는 방안들을 검토한다. 여기서 방역 정책의 대중적ㆍ민주적 토대가 갖는 중요성이 따라 나온다. This article argues that, nearly a year into the pandemic, what is at stake is to make South Korea s response to it more balanced and thus sustainable. What we mean by balanced here is twofold. On the one hand, for a strategy of infectious disease prevention and control to be self-reinforcing, it should not be continued supression nor sudden mitigation; instead, it simultaneously requires one to ramp up ICU capacity and to protect vulnerable population: younger generation in schools and senior citizens living in long-term care facilities in particular. Which consequently indicates where the health care system is consolidated, mitigation also works. On the other, when policymakers feel too much constrained to put a limit on individual freedom and rights, as in the United States or Europe, a sustainable strategy seems not promising. The same holds when these issues are not extensively taken into account nor under deliberation, as in Korea. In this regard, the balanced response matters again. We especially highlight two points. For a more effective strategy of communicating with the general public, we urge policymakers to be active in considering cognitive characteristics of the people as well as the role of their emotion. Furthermore, two specific recommendations, in the form of institutionalized measures, are proposed to ensure that people engage in policy-making relevant to infectious disease prevention and control. Which consequently indicates where democratic legitimacy of the policy is established, its sustainability also works.