• 한국대기환경학회지
• /
• 제32권5호
• /
• pp.513-525
• /
• 2016

To propose an effective ensemble methods in predicting $PM_{10}$ concentration, six experiments were designed by different ensemble average methods (e.g., non-weighted, single weighted, and cluster weighted methods). The single weighted method was calculated the weighted value using both multiple regression analysis and singular value decomposition and the cluster weighted method was estimated the weighted value based on temperature, relative humidity, and wind component using multiple regression analysis. The effects of ensemble average methods were significantly better in weighted average than non-weight. The results of ensemble experiments using weighted average methods were distinguished according to methods calculating the weighted value. The single weighted average method using multiple regression analysis showed the highest accuracy for hourly $PM_{10}$ concentration, and the cluster weighted average method based on relative humidity showed the highest accuracy for daily mean $PM_{10}$ concentration. However, the result of ensemble spread analysis showed better reliability in the single weighted average method than the cluster weighted average method based on relative humidity. Thus, the single weighted average method was the most effective method in this study case.

• Journal of information and communication convergence engineering
• /
• 제3권2호
• /
• pp.67-71
• /
• 2005

Interference occurs by signals received from directions that were different from the signals of the users in a mobile communication system. Various studies have been undertaken, including diversity, equalizer, etc., in order to reduce interference. In this study, the weighted value of the array antenna was obtained to improve signal-to-noise ratio. The weighted value was obtained as an eigen value and an eigen vector by using the correlation coefficient of the signal. The weighted value obtained was then applied to the CDMA system to increase system performance and capacity. Both QPSK and OQPSK modulation systems were applied to analyze performance.

• 한국재료학회지
• /
• 제33권5호
• /
• pp.205-213
• /
• 2023

This paper reports the results of an experimental examination using X-rays to test annealing materials for lapped bearing steel (STB2), to confirm the validity of the weighted averaging analysis method. The distribution behavior for the α_𝜓-sin²𝜓 diagram and the presence or absence of differences in the peak method, half-value breadth method, and centroid method were investigated. When lapping the annealed bearing steel (STB2) material, a residual stress state with a non-directional steep gradient appeared in the surface layer, and it was found that the weighted averaging analysis method was effective. If there is a steep stress gradient, the sin²𝜓 diagram is curved and the diffraction intensity distribution curve becomes asymmetric, resulting in a difference between the peak method, half-value breadth method, and centroid method. This phenomenon was evident when the stress gradient was more than 2~3 kg/mm²/㎛. In this case, if the position of the diffraction line is determined using the centroid method and the weighted averaging analysis method is applied, the stress value on the surface and the stress gradient under the surface can be obtained more accurately. When the stress gradient becomes a problem, since the curvature of the sin²𝜓 diagram appears clearly in the region of sin²𝜓 > 0.5, it is necessary to increase the inclination angle 𝜓 as much as possible. In the case of a lapping layer, a more accurate value can be obtained by considering 𝜎₃ in the weighted averaging analysis method. In an isotropic biaxial residual stress state, the presence or absence of 𝜎₃ can be determined as the presence or absence of strain for sin²𝜓≈0.4.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2002년도 봄 학술발표회 논문집
• /
• pp.369-376
• /
• 2002

The stochastic analysis of semi-infinite domain is presented using the weighted integral method, which is improved to include the higher order terms in expanding the displacement vector. To improve the weighted integral method, the Lagrangian remainder is taken into account in the expansion of the status variable with respect to the mean value of the random variables. In the resulting formulae only the 'proportionality coefficients' are introduced in the resulting equation, therefore no additional computation time and memory requirement is needed. The equations are applied in analyzing the semi-infinite domain. The results obtained by the improved weighted integral method are reasonable and are in good agreement with those of the Monte Carlo simulation. To model the semi-infinite domain, the Bettess's infinite element is adopted, where the theoretical decomposition of the strain-displacement matrix to calculate the deviatoric stiffness of the semi-infinite domains is introduced. The calculated value of mean and the covariance of the displacement are revealed to be larger than those given by the finite domain assumptions which is thought to be rational and should be considered in the design of structures on semi-infinite domains.

• International Journal of Contents
• /
• 제1권2호
• /
• pp.1-4
• /
• 2005

In this paper, we compare 5 weighted distance measures between feature vectors with respect to the recognition performance of the principal component analysis(PCA)-based face recognition method, and propose modified weighted distance. The proposed method was modification of z, the weighted vector. The simulation was performed using the ORL face database, showed the best result for some weighted distances such as weighted manhattan, weighted angle-based, weighted modified manhattan, and weighted modified SSE. We also showed that using some various values of z(weighted values) we can achieve better recognition results that using the existing weighted value.

• Journal of the Korean Statistical Society
• /
• 제30권3호
• /
• pp.467-480
• /
• 2001

This paper considers the use of local $\phi$-divergence measures between posterior distributions under classes of perturbations in order to investigate the inherent robustness of certain classes. The smaller value of the limiting local $\phi$-divergence implies more robustness for the prior or the likelihood. We consider the cases when the likelihood comes form the class of weighted distribution. Two kinds of perturbations are considered for the local sensitivity analysis. In addition, some numerical examples are considered which provide measures of robustness.

• Journal of applied mathematics & informatics
• /
• 제35권1_2호
• /
• pp.191-204
• /
• 2017

A new fifth-order weighted Runge-Kutta algorithm based on heronian mean for solving initial value problem in ordinary differential equations is considered in this paper. Comparisons in terms of numerical accuracy and size of the stability region between new proposed Runge-Kutta(5,5) algorithm, Runge-Kutta (5,5) based on Harmonic Mean, Runge-Kutta(5,5) based on Contra Harmonic Mean and Runge-Kutta(5,5) based on Geometric Mean are carried out as well. The problems, methods and comparison criteria are specified very carefully. Numerical experiments show that the new algorithm performs better than other three methods in solving variety of initial value problems. The error analysis is discussed and stability polynomials and regions have also been presented.

• 한국건설관리학회논문집
• /
• 제16권2호
• /
• pp.46-53
• /
• 2015

한국은행은 2013년 생산자물가지수의 기준연도를 2005년에서 2010년으로 개편하고 기존 고정방식에서 연쇄방식으로 새로운 방식을 적용하였다. 전기공사비지수도 재료분야에서 생산자물가지수를 사용하고 있고, 과거 고정방식 적용으로 현실괴리, 스텝현상 등에 대한 문제점이 지속적으로 제기되어 왔기 때문에, 이에 대한 보완 및 개선방안으로 과거 고정방식에서 현재 연쇄방식으로 변경하였다. 그러나 이러한 편제방식 변경은 가중치 등을 변경시키는 바, 이에 대한 통계적 연속성의 검토가 필요하였다. 따라서 본 연구는 시계열 분석기법을 이용하여 고정방식과 연쇄방식 전기공사비지수의 시계열 특성, 통계특성의 동일성, 양 방식의 대체가능성의 통계연속성 검증을 실시하였다. 그 결과에 의하면, 기존 고정방식에서 현재 연쇄방식을 전기공사비지수에 적용하여도 통계적 연속성이 유지되는 것을 확인할 수 있었다. 따라서 실제 전기공사업 종사자, 전기공사비지수 활용자 등에게 연속적인 시계열 자료로서 인식될 수 있을 것으로 사료된다.

• 응용통계연구
• /
• 제28권1호
• /
• pp.23-31
• /
• 2015

통계적 분석을 할 때 결측치가 발생하는 것은 매우 통상적이다. 이러한 결측치를 대치하는 방법은 여러가지가 있으며, 기존에 사용되는 단일대치법으로 k-nearest neighbor(KNN) 방법이 있다. 하지만 KNN 방법은 k개의 최근접 이웃들 중 극단치나 이상치가 있을 때 편의를 일으킬 수 있다. 본 논문에서는 KNN 방법의 단점을 보완하여 가중 k-최근접이웃(Weighted k-Nearest Neighbors; WKNN) 대치법을 제안하였다. 또한 모의실험을 통해서 기존의 방법과 비교하였다.

• 한국소프트웨어감정평가학회 논문지
• /
• 제15권1호
• /
• pp.25-35
• /
• 2019

우리나라에서 컴퓨터프로그램의 실질적 유사성 여부 판단은 정량적인 유사도를 산출하여 그 결과를 활용하는 방식이 일반적으로 이용된다. 실질적 유사성은 유사한 부분의 양과 질을 고려하여 판단되어야 하는데, 실무에서는 정량적인 유사도 계산 과정에서 가중치를 곱함으로써 유사한 부분의 질을 고려하는 모습을 보인다. 그런데 실질적 유사성 판단과 관련하여 유사한 부분의 양적, 질적인 고려는 동일한 지위에서 순차적으로 이루어져야 한다는 본질적 특징을 고려할 때, 현재와 같은 실무 방식은 적절하다고 할 수 없다. 이에 이와 같은 가중치 활용의 문제를 지적하고, 실질적 유사성 판단을 위한 유사 부분의 질적 평가는 정량적 유사도 판단에 후행하여 그와 동일한 지위에서 이루어져야 함을 제시 및 이를 위한 적절한 실무적 방안을 제언하였다.