• 응용통계연구
• /
• 제18권3호
• /
• pp.661-671
• /
• 2005

본 연구는 반복분할(recursive partitioning)에 의한 군집화 방법을 개발하고 활용 예를 보인다. 노드 분리 기준으로는 Overall R-Square를 채택하였고 실용적인 노드 분리 결정 방법을 제안하였다. 이 방법은 연속형 자료에 대하여 나무 형태의 해석하기 쉬운 단순한 규칙을 제공하면서 동시에 변수선택기능을 제공한다. 환용 예로서 Fisher의 붓꽃데이터와 Telecom 사례에 적용해 보았다. K-평균 군집화와 다른 몇 가지 사항이 관측되었다.

• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2005년도 춘계 학술발표회 논문집
• /
• pp.49-51
• /
• 2005

본 연구는 반복분할(recursive partitioning)에 의한 군집화 방법을 제안하고 활용 예를 제시한다. 이 방법은 나무 형태의 해석하기 쉬운 단순한 규칙을 제공하면서 동시에 변수선택기능을 제공한다.

• Journal of Biosystems Engineering
• /
• 제23권6호
• /
• pp.571-582
• /
• 1998

The PLS(Partial Least Square) and ANN(Artificial Neural Network) were introduced to develop the soluble solids content prediction model of apples which is followed by making a subsequent selection of photosensor. For the optimal PLS model, number of factors needed for spectrum analysis were increased until the convergence of prediction residual error sum of squares. Analysis has shown that even part of the overall wavelength with no pretreatment may turn out better performing. The best PLS model was found in the 800 to 1,100nm wavelength region without pretreatment of second derivation, having $R^2$=0.9236, bias= -0.0198bx, SEP=0.2527bx for unknown samples. On the other hand, for the ANN model the second derivation led to higher performance. On partial range of 800 to 1,100nm wavelengh region, prediction model with second derivation for unknown samples reached $R^2$=0.9177, SEP=0.2903bx in contrast to $R^2$=0.7507, SEP =0.4622bx without pretreatment.

• 재무관리연구
• /
• 제6권2호
• /
• pp.33-69
• /
• 1989

The Purpose of this study is to test empirically the determinants of capital structure of the Korea Listed Firms. In order to accomplish the purpose of this study, both literature survey and empirical test have been made. For the empirical test, agency and asymmetric information factors as well as traditional ones have been throughly reviewed. Traditional factors tested in this study include firm-size, collateral value of the assets, business risk, tax, non-debt tax shields, and industry effects. Agency and asymmetric information factors include growth opportunities, the percentage of outstanding equity held by inside stockholders, and the number of inside stockholders. From the results of the cross-sectional regression analysis, the adjusted R-square is 1931%, and the overall F-value indicates significance. For the analysis period, the signs of the variables except business risk are as predicted. Firm-size, collateral value of the assets, and business risk significant at the.01-.05 level. In order to determine the influence of industry factors on the financial leverage, a total of 8 dummy variables are added to the regression model. The adjusted R-square inclosed by 4.2% for the first analysis period(1983-1985) and 6% for the second analysis period(1986-1987). This suggests that industry factors are significant in explaining the variations in financial leverage across firms. In order to pursue the influence of agency and asymmetric information factors on the financial leverage, again the cross-sectional regression analysis is done for the middle size firms gruop(n=40). The adjusted R-square increased by 9.8% for the first analysis period(1983-1985) and 6.1% for the total analysis period(1983-1987), and all the signs was as predicted. But both the variables except the number of inside stockholders was not significant.

• ETRI Journal
• /
• 제34권4호
• /
• pp.511-517
• /
• 2012

The best linear unbiased estimator (BLUE) is most suitable for practical application and can be determined with knowledge of only the first and second moments of the probability density function. Although the BLUE is an existing algorithm, it is still largely unexplored and has not yet been applied to channel estimation in amplify and forward (AF)-based wireless relay networks (WRNs). In this paper, a BLUE-based algorithm is proposed to estimate the overall channel impulse response between the source and destination of AF strategy-based WRNs. Theoretical mean square error (MSE) performance for the BLUE is derived to show the accuracy of the proposed channel estimation algorithm. In addition, the Cram$\acute{e}$r-Rao lower bound (CRLB) is derived to validate the MSE performance. The proposed BLUE channel estimation algorithm approaches the CRLB as the length of the training sequence and number of relays increases. Further, the BLUE performs better than the linear minimum MSE estimator due to the minimum variance characteristic exhibited by the BLUE, which happens to be a function of signal-to-noise ratio.

• 간호행정학회지
• /
• 제16권1호
• /
• pp.86-100
• /
• 2010

Purpose: This study was done to examine the relationship of job satisfaction and organizational commitment of perioperative nurses. Method: The subjects of this study were 500 perioperative nurses from 11 hospitals. The data were collected by self-reporting questionnaires from Sep. 19 to Sep. 27, 2009. Results: There was statistically significant relationship among the five variables. The analyses of covariance of these five variables revealed overall significant (p<.05). Stepwise linear multiple regression analyses were used to examine the influence of these five variables. Results indicated that the variables for verbal abuse (p<.01), workplace climate (p<.01), internal marketing (p<.001), and job transfer (p<.001) contributed significantly to the job satisfaction (adjusted R square=.426), while the verbal abuse (p<.01), internal marketing (p<.01), leadership style (p<.001) and workplace climate (p<.001) did to the organizational commitment (adjusted R square=.351). Canonical correlation analyses revealed that internal marketing and workplace climate contributed most significantly both to job satisfaction and organizational commitment. Conclusion: This study found that all these five nursing managerial factors were important influential on both job satisfaction and organizational commitment of perioperative nurses. Addressing these factors with further research will surely improve the commitment of these nurses and ultimately lead to better perioperative nursing care.

• Preventive Nutrition and Food Science
• /
• 제18권1호
• /
• pp.67-75
• /
• 2013

Descriptive analysis and consumer acceptance tests were conducted with frozen (FCR), homemade (HCR), and aseptic-packaged (ACR) cooked rice products from two cultivars-IM and SD. FCR was prepared using a rapid freezing process, which may provide consumers with a quality similar to that of HCR. The intensity of the flavors of roasted, glutinous rice, rice cake, and rice starch and the textures of glutinousness, moistness, chunkiness, adhesiveness, and squishiness were all greater in the FCR as compared to the HCR and ACR (p<0.05) in IM and SD cultivars. The differences in sensory characteristics between the FCR and ACR were larger than the equivalent differences between the FCR and HCR. Overall consumer acceptance ratings for FCR in overall aspect, appearance, aroma, and texture were not significantly different compared to those for HCR (p>0.05); however, in most cases these factors showed significant differences when compared with ACR (p<0.05). From partial least square regression analysis, cooked rice was positively related to sweet, transparency, glossiness, roasted, glutinousness, chunkiness, moistness, glutinous rice, adhesiveness, rice shape, rice starch, and squishiness attributes but negatively related to raw rice, old rice, old rice aroma, a particle feeling, off-aroma, white color, scatteredness, slickness, size of cooked rice, and firmness attributes.

• 한국식품영양과학회지
• /
• 제43권1호
• /
• pp.128-134
• /
• 2014

천연 게향을 제조하기 위하여 홍게 농축가공 자숙액($20^{\circ}Brix$)을 원료로 여기에 4종류의 아미노산(proline, glycine, arginine, methionine)과 단당류(fructose)를 선정하여 반응향을 유도하고자 하였다. 4종류의 아미노산과 fructose를 각각 독립변수로 두고 종속변수로는 odor, taste 및 overall acceptance를 두고 중심합성계획법에 따라 반응표면분석법으로 최적 가공조건을 선정하였다. 모델식의 결정계수(R-square)는 odor가 0.88, taste는 0.90, overall acceptance는 0.95였고, 적합결여검증(lack of fit)(P<0.05)에서는 odor(0.01)를 제외하고는 2차식으로의 설계가 유의함을 알 수 있었다(P<0.05). 그러나 odor에서도 결정계수 값이 0.88로서 전체적인 식은 만족할 만하였다. Odor score는 $7.17-0.24[Pro]^2-0.21[Gly]^2-0.21[Arg]^2-0.23[Met]^2-0.27[Fru]^2$의 식으로 나타났으며, 각각 독립변수의 2차항 만이 모두가 유의성(P<0.05)을 나타내었다. 정상점에서의 odor 값은 7.21로 정상점이 최고점이었다. 얻어진 모델식에서의 최적농도는 홍게 농축자숙액($20^{\circ}Brix$)에 proline 0.29 g/100 mL(w/v), glycine 0.63 g/100 mL(w/v), arginine 0.61 g/100 mL(w/v), methionine 0.02 g/100 mL(w/v) 및 fructose 1.07 g/100 mL(w/v)이었으며, 이 조건에서의 실제 관능검사(odor) 값은 7.56이었다.

• 문화기술의 융합
• /
• 제9권4호
• /
• pp.487-494
• /
• 2023

본 연구에서는 무료 딥러닝 도구인 R과 텐서플로우에 대한 성능 비교를 수행하였다. 실험에서는 각 도구를 사용하여 6종류의 심층 신경망을 구축하고 10년간의 한국 온도 데이터셋을 사용하여 신경망을 학습시켰다. 구축된 신경망의 입력층 노드 갯수는 10개, 출력층은 5개로 설정 하였으며, 은닉층은 5, 10, 20개로 설정하여 실험을 진행 하였다. 학습 데이터는 2013년 3월 1일부터 2023년 3월 29일까지 서울시 강남구에서 수집된 온도 데이터 3681건을 사용하였다. 성능 비교를 위해, 학습된 신경망을 사용하여, 5일간의 온도를 예측하고 예측된 값과 실제값을 사용하여 평균 제곱근 오차(root mean square error, RMSE)값을 측정하였다. 실험결과, 은닉층이 1개인 경우, R의 학습 오차는 0.04731176이었으며, 텐서플로우는 0.06677193으로 측정되었으며, 은닉층이 2개인 경우에는 R이 0.04782134, 텐서플로 우는 0.05799060로 측정되었다. 전체적으로 R이 더 우수한 성능을 보였다. 우리는 기계학습을 처음 접하는 사용자들에게 두 도구에 대한 정량적 성능 정보를 제공함으로써, 도구 선택에서 발생하는 어려움을 해소하고자 하였다.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2001년도 춘계학술대회논문집E
• /
• pp.450-455
• /
• 2001

An analytical study is made of transient adjustment process of an initially stationary, stably-stratified fluid in a square. The boundary walls are highly-conducting. The overall Rayleigh number $R_a$ is large. Considerations are given to both opposing (${\delta}w/{\delta}T>0$) and cooperating (${\delta}w/{\delta}T<0$) configurations. The flow character in opposing configuration can be classified into (a) a forced-convection dominaut mode (${\delta}w/{\delta}T>1/\sqrt{2}$), (b) a buoyancy-convection dominant mode ($0<{\delta}w/{\delta}T<1/\sqrt{2}$), and (c) a static mode (${\delta}w/{\delta}T{\cong}1/\sqrt{2}$). Global evolutionary processes are depicted. and physical rationalizations are provided.