• Journal of the Earthquake Engineering Society of Korea
• /
• v.8 no.2
• /
• pp.35-44
• /
• 2004

In the capacity spectrum method(CSM) using a linear response spectrum, the peak response of an inelastic system under a given earthquake load is estimated transforming the system into the equivalent elastic one. The CSM for estimating the peak inelastic response is evaluated in this paper. The equivalent period and damping ratio are calculated using the ATC-40, G lkan, Kowalsky, and Iwan methods, and the performance points are obtained according to the procedure B of ATC-40. Analysis results indicate that the ATC-40 method generally underestimates the peak response resulting in the unsafe design, while the G lkan and Kowalsky methods overestimate the responses. The Iwan method produces the values between those by the ATC-40 method and the G lkan and Kowalsky methods, and estimates the responses relatively closer to the exact ones. Further, it is found that the Kowalsky method gives the negative equivalent damping ratios depending on the hardening ratios, and thereby can not be used to estimate the responses in some cases.

• The Korean Journal of Applied Statistics
• /
• v.27 no.6
• /
• pp.1003-1016
• /
• 2014

Proper missing data imputation is an important procedure to obtain superior results for data analysis based on survey data. This paper deals with both a model based imputation method and model estimation method. We utilized a Bayesian method to solve a boundary solution problem in which we applied a maximum likelihood estimation method. We also deal with a missing mechanism model selection problem using forecasting results and a comparison between model accuracies. We utilized MWPE(modified within precinct error) (Bautista et al., 2007) to measure prediction correctness. We applied proposed ML and Bayesian methods to the Korean presidential election exit poll data of 2012. Based on the analysis, the results under the missing at random mechanism showed superior prediction results than under the missing not at random mechanism.

• Journal of Korean Society of Transportation
• /
• v.26 no.2
• /
• pp.81-88
• /
• 2008

This paper presents the travel demand estimation using interval estimation methods during the trip generation stage, and then followed the other three stages of the four stage trip estimation. We have used real data of Dae-jun City. To estimate travel demand using the interval estimation method, a reliability level was set to 95% by a upper bound value, a middle value and a lower bound value. The four stage traffic demand analysis procedure was equally applied and finally interval traffic was estimated. The result showed a difference between maximum values and middle values depending on the destination during the trip generation stage. It depends on an explanation ability of regression analysis. Most of interval estimation ratio resulted in the traffic assignment stage showed ${\pm}5{\sim}18%$ difference on the average and ${\pm}30{\sim}50%$ at the most.

• Tunnel and Underground Space
• /
• v.14 no.5
• /
• pp.353-362
• /
• 2004

Prediction of fragmentation in bench blasting is one of the most important factors to establish the production plan. It is widely accepted that fragmentation could be accurately predicted using the Kuz-Ram model in bench blasting. Nevertheless, the model has an ambiguous or subjective aspect in evaluating the model parameters such as joint condition, rock strength, density, burden, explosive strength and spacing. This study proposes a new method to evaluate the parameters of Kuz-Ram model, and the predicted mean fragment sizes using the proposed method are examined by comparing the measured sizes in the field. The results show that the predictions using Kuz-Ram model with the proposed method coincide with field measurements, but Kuz-Ram model does not reflect the in-situ rock condition and hence needs to be improved.

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.4
• /
• pp.163-173
• /
• 2018

This paper performed the prediction of the acoustic loads applied to the surface of the flight vehicle during flight. Acoustic loads during flight arise from the pressure fluctuations on the surface of body. The conventional method of predicting the acoustic loads in flight uses semi-empirical method derived from theoretical and experimental results. However, there is a limit in obtaining the flow characteristics and the boundary layer parameters of the flight vehicle which are used as the input values of the empirical equation through experiments. Therefore, in this paper, we use the hybrid method which combines the results of CFD (Computational Fluid Dynamics) with semi-empirical methods to predict the acoustic loads acting on flight vehicle during flight. For the flight vehicle with cone-cylinder-flare shape, acoustic loads were estimated for the subsonic, transonic, supersonic, and Max-q (Maximum dynamic pressure) condition flight. For the hybrid method, two kind of boundary layer edge estimation methods based on CFD results are compared and the acoustic loads prediction results were compared according to empirical equations presented by various researchers.

• Journal of Korea Water Resources Association
• /
• v.33 no.S1
• /
• pp.788-793
• /
• 2000

본 연구는 k서로 특성이 다른 8개의 도시를 검토 대상지역으로 선정하여 사용수량의 추정방법중 이론 곡선법을 이용하는 경우의 포화값 K의 영향을 비교 검토한 것이다. 이 연구결과는 상수사용량을 예측할 때 일어날 수 있는 오류를 최소한으로 줄이고 해당도시의 예측값을 결정할 때 보다 합리적으로 접근하는데 기초자료가 될 수 있을 것이다.

• The KIPS Transactions:PartB
• /
• v.8B no.3
• /
• pp.231-236
• /
• 2001

일반적인 다층 신경망에서 가중치의 갱신 알고리즘으로 사용하는 오류 역전과 방식은 가중치 갱신 결과를 고정된(fixed) 한 개의 값으로 결정한다. 이는 여러 갱신의 가능성을 오직 한 개의 값으로 고정하기 때문에 다양한 가능성들을 모두 수용하지 못하는 면이 있다. 하지만 모든 가능성을 확률적 분포로 표현하는 갱신 알고리즘을 도입하면 이런 문제는 해결된다. 이러한 알고리즘을 사용한 베이지안 신경망 모형(Bayesian Neural Networks Models)은 주어진 입력값(Input)에 대해 블랙 박스(Black-Box)와같은 신경망 구조의 각 층(Layer)을 거친 출력값(Out put)을 계산한다. 이 때 주어진 입력 데이터에 대한 결과의 예측값은 사후분포(posterior distribution)의 기댓값(mean)에 의해 계산할 수 있다. 주어진 사전분포(prior distribution)와 학습데이터에 의한 우도함수(likelihood functions)에 의해 계산한 사후확률의 함수는 매우 복잡한 구조를 가짐으로 기댓값의 적분계산에 대한 어려움이 발생한다. 따라서 수치해석적인 방법보다는 확률적 추정에 의한 근사 방법인 몬테 칼로 시뮬레이션을 이용할 수 있다. 이러한 방법으로서 Hybrid Monte Carlo 알고리즘은 좋은 결과를 제공하여준다(Neal 1996). 본 논문에서는 Hybrid Monte Carlo 알고리즘을 적용한 신경망이 기존의 CHAID, CART 그리고 QUEST와 같은 여러 가지 분류 알고리즘에 비해서 우수한 결과를 제공하는 것을 나타내고 있다.

• The Korean Journal of Financial Management
• /
• v.14 no.2
• /
• pp.229-252
• /
• 1997

본 연구에서는 실무 및 학계에 종사하는 45명의 전문가 집단을 대상으로 쌍별비교(pairwise comparision)에 의한 설문조사에서 얻어진 전문가들의 의견을 AHP 분석을 통하여 종합하는 과정을 거쳐 부도예측모형을 설계하여 검증한 뒤, LOGIT모형과 비교하였다. 본 연구에 의하면 부도예측모형에서 정량적인 정보보다 정성적인 정보가 더 중요한 역할을 한다는 D.Bunn-G.Wright(1991)의 연구와 일치하는 결과를 얻을 수 있었다. 본 연구에서 발견된 분석결과를 요약하면 다음과 같다. 첫째로 LOGIT 모형과 AHP 모형에서 모두 정량적인 정보만을 고려하는 경우보다 정성적인 정보를 함께 고려한 모형에서 부도예측율이 더 높은 것으로 나타나고 있어 부도가능성을 예측하는데 있어 정성적인 정보가 중요한 역할을 한다는 결론을 얻었다. 둘째로 AHP를 이용한 부도예측 모형을 설계할 때 각 속성에 대한 전문가(45명)들의 의견을 종합하는 방법으로 산술평균과 기하평균을 이용한 검증결과에 의하면 기하평균방법을 통하여 전문가들의 의견을 종합하는 것이 보다 합리적이라는 실증적 증거를 얻을 수 있었다. 셋째로 Akaike의 기준값을 분석한 결과에 의하면 LOGIT 모형은 정량적인 정보와 정성적인 정보를 모두 이용한 모형이 가장 우수한 것으로 판명되었고, 모형의 부도예측력도 가장 높은 것으로 밝혀졌다. AHP 모형은 정성적인 정보만을 이용한 모형에서 가장 높은 부도예측을을 나타내었으며, 기하평균을 이용한 AHP 모형은 LOGIT 모형보다 항상 높은 부도예측율을 보여주었다.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.6
• /
• pp.1601-1607
• /
• 2016

Recently, weather information has been increasingly used in various area. This study presents the necessity of hourly weather information for electricity demand forecasting through correlation analysis and multivariate regression model. Hourly weather data were collected by Meteorological Administration. Using electricity demand data, we considered TBATS exponential smoothing model with a sliding window method in order to forecast electricity demand. In this paper, we have shown that the incorporation of weather infromation into electrocity demand models can significantly enhance a forecasting capability.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.7
• /
• pp.11-18
• /
• 2023

In this paper, we evaluate deep learning time series forecasting models. Recent studies show that those models perform better than the traditional prediction model such as ARIMA. Among them, recurrent neural networks to store previous information in the hidden layer are one of the prediction models. In order to solve the gradient vanishing problem in the network, LSTM is used with small memory inside the recurrent neural network along with BI-LSTM in which the hidden layer is added in the reverse direction of the data flow. In this paper, we compared the performance of Informer by comparing with other models (LSTM, BI-LSTM, and Transformer) for real Nitrogen dioxide (NO2) data. In order to evaluate the accuracy of each method, mean square root error and mean absolute error between the real value and the predicted value were obtained. Consequently, Informer has improved prediction accuracy compared with other methods.