http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Education and Economic Growth: an OECD Panel Study
이재우,문성아 한국자료분석학회 2009 Journal of the Korean Data Analysis Society Vol.11 No.3
In this study we empirically test the hypothesis that education matters on economic growth. Based on theories of human capital development, we postulate that education leads to economic growth. Education expenditures and enrollment rates in primary, secondary, and tertiary educations are especially chosen as independent variables to test if they have an impact on economic growth, or per capita GDP in our regression. To consider the hetero- skedasticity of error terms, we turn to GLS regression to see the effects of education factors on economic growth. In addition, we divide the OECD panel data into two subsets; Group A as a relatively high income country group and Group B as a low income group. Overall, education variables turn out to be significant in explaining per capita GDP growth. In group A, or relatively wealthy group, enrollment in tertiary education sector tend to be significantly important, reflecting that educational effects on economic growth tend to be even stronger in highly knowledge-driven countries in OECD.
EEG 신호를 이용한 보행 보조 로봇 이용자의 의도인지 방법
이재우 대한기계학회 2019 大韓機械學會論文集A Vol.43 No.12
This paper presents a mathematical modeling that can describe the characteristics of personal neuro-activities based on the EEG signal captured from the EEG device on the surface of scalp. In case of elderly people, walking assistant device is used to help walking. In order to drive electric motors that are attached to the walking assistant device using brain wave, interpretation of brain wave is needed. Brain waves are extracted as a form of EEG signal through the headset of Neurosky’s MindWave. EEG signals are divided into alpha, beta, gamma, delta and theta to make 8 dimensional vectors. As dimension of output signals are too high, characteristics is difficult to understand, reduction of dimension is carried out using principal component analysis method. First, feature is extracted from the experimental data using PCA and mathematical model is derived and verified. Second, using this model, classification tests is carried out. KNN for classification are applied on the experimental data. Final results show well that the intent of human being can be discriminated based on proposed method. 이 논문에서는 두피 위에 장착된 EEG 장비로부터 취득한 EEG 신호에 근거하여 사람의 뇌 활동의 특성을 서술할 수 있는 수학적 모델 방법이 제시된다. 노약자의 경우 보행을 위해 보행보조장치를 사용한다. 보행보조장치에 전동모터를 부착하고 보행자의 뇌파 신호를 이용하여 전동모터를 동작시키기 위해서는 보행자의 의도를 나타내는 뇌파의 해석이 필요하다. 뇌파는 EEG 신호의 형태로 헤드셋기기를 이용하여 추출한다. EEG 신호는 1초에 한 번씩 알파, 베타, 감마, 델타, 쎄타 등의 주파수별 신호로 나뉘어 8차원 벡터로 구성된다. 출력신호의 차원이 너무 높아 특징을 파악하기 어려우므로 주성분 분석(PCA) 방법으로 차원축소를 시도한다. 먼저 신호로부터 특징이 주성분 분석법(PCA)을 이용하여 실험 데이터로부터 추출된다. 두 번째, 제시된 모델을 사용하여 분류작업이 수행된다. 분류를 위해 KNN(K-Nearest Neighbor) 방법이 실험데이터에 적용된다. 최종적인 실험 결과는 인간의 의도가 제안된 방법을 사용하여 분별 될 수 있음을 나타낸다.
Scaling Behaviors of Plant-Pollinator Mutualistic Networks
이재우,이경은,맹성은,황준경 한국물리학회 2008 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.53 No.6
We consider three plant-pollinator mutualistic networks, including a montane forest, a beech forest and a meadow. The plant and the pollinator connect by links of interacting strengths between species. We obtained cumulative distribution functions (CDFs) of the degree and the strength for the networks. The CDFs of the degree for all species, pollinators and plant, follow a power law or a stretched exponential function. The CDF of the degree for a meadow food web, MEM food web (Mommett 1999), is well fitted by a stretched exponential function. However, the CDF of the degree for the montane forest food web, INO food web (Inouye and Pkye 1998) and for the beech forest food web, KAT food web (Kato et al. 1990) show a power law. The CDFs of the strength also follow a power law, except for the CDF of the pollinator for the KAT food web. The CDFs of the degree and the strength for mutualistic networks depend on the types and the locations of the food webs.