http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
오늘 본 자료
김은미 한국감성과학회 2015 춘계학술대회 Vol.2015 No.-
우리나라는 급속하게 고령사회로 편입되면서 노인 우울이 심각한 사회문제가 되고 있다. 노인 우울은 치매 가능성과 사망률 증가에 영향을 미치므로 조기 진단과 다양한 처방이 시급하다고 할 수 있다. 우울증은 정상인에 비해 긍정 정서는 짧게, 부정 정서는 더 강하게 오랫동안 경험함으로써 회복탄력성이 낮아진다(Henriques & Davidson, 1991; Davidson, 2000). 긍정 및 부정 정서는 각각 좌뇌의 접근적 행동 성향(BAS)과 우뇌의 회피적 행동 성향(BIS)과 연관되는데, 우울증에서는 좌뇌 비활성 혹은 우뇌 과활성이 나타날 수 있다(Davidson, Jackson, & Kalin, 2000). 좌뇌 활성인 사람들은 동일한 중재법이라도 더 큰 효과를 보이는데, 이는 긍정적이며 적극적인 태도 때문일 수 있다(김은미, 2013). 전두엽 알파 비대칭은 우울증과 정상인의 차이를 보여주는 상태 (trait) 지표일 가능성이 제시되어 왔다(최승원, 제갈은주, 안창일, 2008; Stewart, Bismark, Towers, et al., 2010). 우울증의 다양한 하위 증상이 각 주파수 대역 비대칭과 연관될 수 있으므로, 알파 비대칭 이외에도 베타 비대칭 또는 세타 비대칭과 우울증간의 관련성에 대한 가능성도 제시되고 있다(정규희, 이준석, 이장한, 2008). 본 연구에서는 임상적 우울증 환자가 아닌 일반 노인의 우울 수준과 전전두엽 알파 비대칭 및 베타 이어느 정도 관련이 있는지 분석해 보고자 한다.
온라인 P2P 환경에서 딥러닝을 적용한 다분류 기반 개인신용등급 예측모형
김은미,박지영 한국인터넷전자상거래학회 2019 인터넷전자상거래연구 Vol.19 No.4
In this study, we propose a multi–class personal credit rating prediction model applying deep learning to accurately predict personal credit rating as a way to reduce investor's investment risk. In P2P lending, credit information on borrowers has an important influence on investment decisions, since investors make investment decisions based on objective information of borrowers without special collateral. To develop the multi-class prediction model, we use the Lending Club loan data to classify the customer's grade into four classes, and randomly extract 1,000 data from each grade and use a total of 4,000 data. We apply the method which is MDA, ANN, one-against-one, Weston and Watkins, Crammer and Singer, and DNN. The MDA and ANN were used as benchmarking models. As the results, DNN shows higher prediction performance than other models, so that deep learning can be applied with excellent performance in multi-class problems. In addition, the confusion matrix shows that the DNN model can reduce the investment risk of investors rather than the multi-class SVM model. By applying the DNN, we can make up for the defect of the multi-class SVM, which is obtained through multiple binary classifications, and can be applied more efficiently in big data environment. The proposed method can protect investors from investment risk by providing more detailed information on investment decisions in online P2P finance.
International Political Economy on the GATT/WTO dispute participation : US dispute initiation
김은미 한국관세학회 2012 관세학회지 Vol.13 No.4
This study investigates how the US government makes decision for filing complaint to GATT/WTO based on the four international political economic theories of Frieden and Lake (2009). To do that this study use regression analysis, applying it to 20152 pooling data (four-digit SIC industries in the manufacturing sector) for the years 1948 through 2006. This study generate a previously unused dependent variable, Propensity to File Complaint (PFC) which is measured by the number of consultation requested by the US, with independent variable that represent each theoretical perspective. The empirical findings indicate that the “domestic institutional perspective” and “domestic political perspective” dominate the analysis. Specifically partisan control of the executive and size and the level of organization of interest groups appear highly relevant for understanding US dispute participation decision.
국제 제품 전략과 수출 성과: 혁신 역량의 조절 효과를 중심으로
김은미 한국무역통상학회 2019 무역통상학회지 Vol.19 No.4
This study employs the resource-based view and the organizational learning perspective to understand how cost leadership and differentiation strategies influence on export performance and the distinctive roles of innovation capabilities on these links between international product strategies and export performance relatively. Using survey data from 216 Korean exporting firms, the findings indicate that cost leadership and differentiation strategies enhance export performance. Moreover, while exploitative innovation capability strengthens the relationship between cost leadership strategy and export performance, exploratory innovation capability enhances the link between differentiation strategy and export performance. This study suggests theoretical background for strategic fits between international product strategy and innovation capability to enhance export performance. Furthermore, the findings underscore needs for managers to invest in a certain type of international product strategy to ensure the development of innovation capability differently.