• Journal of the Korea Society of Computer and Information
• /
• v.16 no.8
• /
• pp.39-47
• /
• 2011

Ergezinger's method, one of the layer-by-layer algorithms used for multilyer perceptrons, consists of an output node and can make premature saturations in the output's weight because of using linear least squared method in the output layer. These saturations are obstacles to learning time and covergence. Therefore, this paper expands Ergezinger's method to be able to use an output vector instead of an output node and introduces a learning rate to improve learning time and convergence. The learning rate is a variable rate that reflects the correlation coefficient between new weight and previous weight while updating hidden's weight. To compare the proposed method with Ergezinger's method, we tested iris recognition and nonlinear approximation. It was found that the proposed method showed better results than Ergezinger's method in learning convergence. In the CPU time considering correlation coefficient computation, the proposed method saved about 35% time than the previous method.

• Journal of Korean Association for Spatial Structures
• /
• v.8 no.1
• /
• pp.57-67
• /
• 2008

In general, the response spectrum analysis(RSA) method is wifely used for seismic analysis of building structures, and the time history analysis(THA) is applied for computation of structural vibration caused by equipments, machines and moving loads, etc. However, compared with the RSA method, the THA method is very complex, difficult and time consuming. In this study, the maximum responses for the vertical vibration are calculated conveniently by the RSA method. At first, the process for the RSA in excitation is proposed, and the maximum modal responses are combined by CQC and SRSS methods. Also, the responses obtained by the two modal combination methods are compared to the responses by the THA. And the correlation coefficients for human activities is proposed, and the RSA responses obtained by used to the correlation coefficients are calculated. Finally, results of the proposed method are compared with those of the time history analysis and correlation coefficients should be considered for the RSA of floor structure subjected to group dynamic loads.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.211-211
• /
• 2018

미해양대기청 기후예측센터(Climate Prediction Center, NOAA)에서 제공하고 있는 기후지수(climate indices)를 예측인자로 하고 금강유역의 5~6월의 강우량을 예측대상으로 하는 원격상관기반 통계모형을 구축하였다. 1988년부터 2017년까지의 30년 자료에 대해 예측인자와 예측대상간의 시간지연상관분석을 수행한 결과 NAO(North Atlantic Oscillation), EP/NP(East Pacific/North Pacific Oscillation), EA(East Atlantic Pattern), WP(Western Pacific Index) 등과 상관성이 높은 것으로 분석되었으며, 이러한 시간지연 기후지수를 이용하여 4개월전에 5,6월 강수량을 예측할 수 있는 다중회귀모형을 개발하였다. 관측 강우량 아노말리가 큰 경우에는 다소 과소 예측되고, 아노말리가 작은 경우에는 다소 과다 예측되는 경향을 보였지만 관측 강우량과 예측 강우량간의 상관계수가 0.75로서 비교적 우수한 예측 결과를 나타내었다. 5~6월 강우량 아노말리의 3분위 예측성을 평가한 결과 평년이상 적중률은 77.8%, 평년수준은 81.8%로서 예측 성공률이 높았으며, 5, 6월 누적강우량이 매우 작았던 92년과 95년을 제외하고는 강우량이 적은 해에도 예측성이 우수하여 평년이하 적중률이 70.0%를 나타내었다. 따라서 본 개발모형은 최소 4개월 이전 선행시간을 가지고 늦봄, 초여름강우량을 예측할 수 있는 저비용의 가뭄 예측 도구로 유용하게 활용될 수 있을 것이다.

• Korean Journal of Environment and Ecology
• /
• v.10 no.1
• /
• pp.49-57
• /
• 1996

To figure out the relation between the shade hours and the landscape tree growth in the apartment housing areas, the present sizes and planting positions of 4 tree species in Gwacheon-si apartment housing areas were surveyed. Then, shade hours were analyzed and the data were analyzed by simple linear regression method. As a whole, the R$^{2}$ was too low to generalize the regression equation. Therefore, it was presumed that the gravity of shade hours in landscape tree growth in this sample site was relatively lower than that of any other environmental factors. However, it was presumed that the characteristics of shade intolerant and tolerant tree were turned up, because Pinus strobus showed a low negative correlation with shade housm and Acer palmatum and Magnolia denudata showed a low positive correlation with shade hours generally. And, it was proved that the statistically significant cases were the tree diameter at root collar and tree sidth of Acer palmatum and tree width of Magnolia denudata with shade hours showing a low correlation coefficient less than 0.4.

• Proceedings of the KSPS conference
• /
• 1996.02a
• /
• pp.125-135
• /
• 1996

본 논문에서는 자연스러운 음성을 합성하기 위하여, 한국어 음운지속시간의 변화에 있어서 문절과 구내의 음절수와 음절의 위치에 의한 영향과 인접하는 음운의 영향에 대하여 통계적으로 분석하였고, 분석된 시간 특징을 제어 요소로 하는 회귀트리를 생성하여 음운 지속시간을 모델 화하였다. 또한, 제안된 음운 지속시간 모델에 의해 예측실험을 행하여, 측정치와 예측치간의 다중 상관계수가 0.74정도이고, 각 음운의 예측오차의 75%이상이 25ms이내로 제안된 모델의 타당성이 입증되었다.

• Journal of Korean Society of Transportation
• /
• v.17 no.3
• /
• pp.47-60
• /
• 1999

In this study, the dynamic route choice behavior of driver is assumed to be affected by the current conditions of traffic environments as well as the Past traffic environments and activities. The repeated survey of multi-timed to owner drove in Chon-ju city by the virtual traffic information system was performed by the stated preference method. And the LISREL(An analysis of linear structural relationship) model was used. As the results. the variable Parameter and t-value of travel time information on applied model was high and their results have an effect greatly to the route choice. After all, the route choice was negative from Kirin-ro, which Penetrates the downtown, and positive from Chunbyun-ro which was a roundabout way was confirmed. Estimated result of the reasonability, the highest suitable model was ode which applied the concept of serial correlation and stated dependence and was shown applying rightfulness to dynamic model. As the serial correlation and stated dependence parameter value, when time interval was large, parameter value was small and the serial correlation and stated dependence was in inverse proportion ratio to the time interval.

• Journal of Korea Water Resources Association
• /
• v.53 no.7
• /
• pp.533-543
• /
• 2020

In order to provide information for proper management of groundwater resources, it is necessary to estimate the rise time of groundwater level by calculating the delay time between the time series of precipitation and groundwater level and to understand the characteristics of groundwater level variation. In this study, total delay time (TDT) and cross correlation coefficient between the moving averaged precipitation generated by using the moving average method to take into account the preceding precipitation and the groundwater level were calculated and analyzed for the nine groundwater level monitoring wells in the Pyoseon watershed in the southeast of Jeju Island. As a result, when the moving averaged precipitation was used, the correlation with the groundwater level was higher in all monitoring wells than in the case of using the raw precipitation, so that it was possible to more clearly estimate the delay time between precipitation and groundwater level. When using the moving averaged precipitation, it had cross correlation coefficients of up to 0.57 ~ 0.58 with the time series data of the groundwater level, and had a relatively high correlation when considering the preceding precipitation of about 24 days on average. The TDT was about 32 days on average, and it was confirmed that the consideration of preceding precipitation plays an important role in estimating the TDT because the days of moving averaged precipitation greatly influences the calculation of the TDT. In addition, through the use of moving averaged precipitation, we found an error in estimating the TDT due to the use of raw precipitation. Through the method of estimating the TDT used in this study and the use of the R code for estimating the TDT presented in the appendix of this paper, it will be possible to estimate the TDT for other regions in the future relatively easily.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.307-307
• /
• 2011

ARF(Areal Reduction Factor, 면적우량감소계수)는 지점강우량을 면적 평균 강우량으로 변환하는 환산계수로 정의되며, 유역의 지형학적 특성과 강우의 공간적 분포특성을 반영한 유역단위의 ARF의 개발이 요구된다. 하지만 국내의 ARF는 대부분 한강유역을 대상으로 하고 있어 한강유역과 지형학적, 수문 기상학적 특징이 상이한 유역에 대하여 연구 결과를 적용하기는 많은 제약이 따를 것으로 판단된다. 따라서 본 연구에서는 남강댐 유역의 ARF를 산정하기 위해 7개의 강우관측소(산청, 삼가, 신안, 안의, 운봉, 태수, 함양)로부터 시강우자료(1990년~2010년)를 수집한 후 14개의 재현기간, 6개의 지속시간에 대한 지수형 ARF 회귀식을 산정하였다. 그 결과 남강댐 유역의 지수형 ARF 회귀식의 결정계수는 0.80~0.99로 높은 상관성을 나타내었다. 그리고 남강댐 유역의 ARF와 첨두홍수량의 관계를 분석하기 위해 남강댐 유역내의 산청유역을 대상으로 재현기간 100년, 지속시간 24시간에 대한 홍수량을 모의하였다. 그 결과 ARF의 적용 전 후의 첨두홍수량은 10% 이상 감소하는 것으로 나타났다. 따라서 남강댐 유역의 기상학적 특성을 고려한 첨두홍수량 산정을 위해서는 본 연구에서 제안한 ARF 회귀식이 유용할 것으로 판단된다.

• Journal of Korea Water Resources Association
• /
• v.49 no.4
• /
• pp.327-333
• /
• 2016

Due to climate changes accelerated by global warming, South Korea has experienced regional climate variations as well as increasing severities and frequencies of extreme weather. The precipitation in South Korea during the summer season in 2013 was concentrated mainly in the central region; the maximum number of rainy days were recorded in the central region while the southern region had the minimum number of rainy days. As a result, much attention has been paid to the importance of flood control due to damage caused by spatiotemporal intensive rainfalls. In this study, forecast rainfall data was used for rapid responses to prevent disasters during flood seasons. For this purpose, the applicability of numerical weather forecast data was analyzed using the ground observation rainfall and inflow rate. Correlation coefficient, maximum rainfall intensity percent error and total rainfall percent error were used for the quantitative comparison of ground observation rainfall data. In addition, correlation coefficient, Nash-Sutcliffe efficiency coefficient, and standardized RMSE were used for the quantitative comparison of inflow rate. As a result of the simulation, the correlation coefficient up to six hours was 0.7 or higher, indicating a high correlation. Furthermore, the Nash-Sutcliffe efficiency coefficient was positive until six hours, confirming the applicability of forecast rainfall.

• Proceedings of the KIPE Conference
• /
• 2016.07a
• /
• pp.15-16
• /
• 2016

본 논문에서는 케이블 결함과 위치를 검출하기 위한 SSTDR(spread spectrum domain reflectomerty) 기반의 진단 장치를 개발하였다. 개발된 진단 장치는 sstdr 기법의 검출 성능을 향상시키기 위해 시간-주파수 상관 분석을 이용하여 기준신호(reference signal)의 상관계수 최댓값 위치를 검출 한 후 기준신호를 제거하여 반사신호의 상관계수의 최댓값 위치를 검출하는 2단계 과정을 갖는다. 개발된 케이블 결함 진단 장치는 실증 시험장을 구축하고, 성능시험을 실시한 결과, ${\pm}1%$ 내 외의 오차를 갖는 것으로 확인되었다.