• The Korean Journal of Applied Statistics
• /
• v.17 no.3
• /
• pp.459-473
• /
• 2004

It has well known that an exhaustive search algorithm suggested by Breiman et. a1.(1984) has a trend to select the variable having relatively many possible splits as an splitting rule. We propose an algorithm to overcome this variable selection bias problem and then construct unbiased regression trees based on the algorithm. The proposed algorithm runs two steps of selecting a split variable and determining a split rule for binary split based on the split variable. Simulation studies were performed to compare the proposed algorithm with Breiman et a1.(1984)'s CART(Classification and Regression Tree) in terms of degree of variable selection bias, variable selection power, and MSE(Mean Squared Error). Also, we illustrate the proposed algorithm with real data sets.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.2
• /
• pp.237-245
• /
• 2012

This paper describes an acoustic echo canceller with double-talk by the blind signal separation. The acoustic echo canceller is deteriorated or diverged in the double-talk period. So we use the blind signal separation to estimate the near-end speech signal and to eliminate the estimated signal from the residual signal. The blind signal separation extracts the near-end signal with dual microphones by the iterative computations using the 2nd order statistical character. Because the mixture model of blind signal separation is multi-channel in the closed reverberation environment, we used the copied coefficients of echo canceller without computing the separation coefficients. By this method, the acoustic echo canceller operates irrespective of double-talking. We verified performances of the proposed acoustic echo canceller by simulations. The results show that the acoustic echo canceller with this algorithm detects the double-talk periods thoroughly, and then operates stably in the normal state without the divergence of coefficients after ending the double-talking. And it shows the ERLE of averagely 20dB higher than the normal LMS algorithm.

• Environmental and Resource Economics Review
• /
• v.31 no.4
• /
• pp.849-864
• /
• 2022

In case of a specific sector being divided into sub-sectors, this study presents a process for estimating an input-output table, which is frequently used as basic data in fields of energy and environment economics. RAS method, which is universally used for this case, requires information on production, intermediate input sum, and intermediate demand sum for each sector in the new table. But in many cases, it is difficult to secure information on intermediate demand sum by sector. This study suggests a process for estimating a new input-output table without using information of intermediate demand sum in the case of sector separation, under the assumption that information of production value and intermediate input sum by sector are available. The key idea is that the values of many elements in the input-output table after disaggregation are the same as those in the table before disaggregation and that the sum of the elements after disaggregation, equals the values of the elements before disaggregation. The process of estimating the intemediate transaction matrix or the input coefficient matrix is presented by using these information instead of intermediate demand sum information. A small-scale simulation shows that the average error rate of the process proposed in this study is about 11.23% in estimating input coefficients, which is smaller than the 11.30% estimation error of RAS using the information of intermediate demand sum. However, since it is known in the literature that using additional information does not always improve estimation performance compared to not using it, additional research on various simulations is needed to apply the method of this study to reality.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.4
• /
• pp.210-216
• /
• 2005

We Propose a new feature normalization scheme based on eigenspace for achieving robust speech recognition. In general, mean and variance normalization (MVN) is Performed in cepstral domain. However, another MVN approach using eigenspace was recently introduced. in that the eigenspace normalization Procedure Performs normalization in a single eigenspace. This Procedure consists of linear PCA matrix feature transformation followed by mean and variance normalization of the transformed cepstral feature. In this method. 39 dimensional feature distribution is represented using only a single eigenspace. However it is observed to be insufficient to represent all data distribution using only a sin91e eigenvector. For more specific representation. we apply unique na independent eigenspaces to cepstra, delta and delta-delta cepstra respectively in this Paper. We also normalize training data in eigenspace and get the model from the normalized training data. Finally. a feature space rotation procedure is introduced to reduce the mismatch of training and test data distribution in noisy condition. As a result, we obtained a substantial recognition improvement over the basic eigenspace normalization.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.823-825
• /
• 2005

컴퓨터 비전 연구에서의 주요 관심은 객체의 특징을 이용하여 객체를 분간하거나 또는 계수하는데 있어 왔다. 최근 대단위의 사람들이 운집하는 공공장소에서의 사고에 대비한 대책의 기준으로 혼잡도라는 점보의 중요성이 대두되고 있다. 본 실험에서는 객체들이 존재하는 전경(Foreground) 영역을 객체들이 없는 배경 영역(back ground)으로부터 분리한 후 전경 영역에서의 edge pixel 들의 수를 계수하여 혼잡도의 정도를 구한다. 전경 영역과 배경 영역은 소 영역별로 RGB에 대한 표준편차와 평균을 비교 분석해서 구분하고 배경 영역을 삭제한다. 전경 영역에서 edge detection 방법을 이용하여 환경에 알맞은 edge pixels수를 계수하고 pixels수와 혼잡도 사이의 관계를 구한다. 이러한 측정 방법의 장점은 다양한 환경에서도 혼잡도라는 기본 특징정보를 추출할 수 있다는 것이다.

• The Korean Journal of Applied Statistics
• /
• v.12 no.2
• /
• pp.573-584
• /
• 1999

본 연구는 이변량 회귀모형을 이변량 자료에 적용할 때 이변량 자료(분리형 자료) 이외에 이변량 자료를 합산한 일변량 자료(통합형 자료)를 동시에 사용하는 문제를 고찰하였다. 특징을 파악하기 위하여 설명변수가 하나인 경우를 다루었는데 통합형 자료의 첨가효과를 회귀계수의 추정량의 평균제곱오차의 크기로서 측정하면서 효과와 이변량 모형과의 관계를 조사하였다. 최대우도 추장량의 특성으로부터 대표본의 성질을 추출하고 또한 모의실험을 통하여 소표본에서도 대표본의 성질이 만족하는지 조사하였고 끝으로 실제 자료에 적용하여 보았다.

• Proceedings of the Korean Environmental Sciences Society Conference
• /
• 2005.11a
• /
• pp.29-32
• /
• 2005

주거지역 A와 주거지역 B의 PM10 평균농도 비교에서 산업지역과 인접한 주거지역 A에서 계절에 관계없이 26-32% 높은 농도를 나타내어 산업지역에 인접한 주거지역 A가 먼 주거지역 B 보다 PM10의 영향에 더 크게 노출된 것을 확인 할 수 있었다. PM10에 포함된 Cd, Cr, Cu, Fe, Mn, Ni, Pb, Zn 등의 무기성분은 대부분 인위적인 발생원과 관련성이 있는 대기 오염물질로, 인위적인 배출원의 영향정도를 파악하기 위하여 풍성계수 분석결과 높은 풍성계수를 나타내어 인위적인 오염원임을 확인 할 수 있었고, 그 평균농도는 주거지역 A가 주거지역 B 보다 높게 나타나 주거지역 A가 주거지역 B 보다 산업 오염원의 노출에 더 많은 영향을 받는 다는 것을 확인 할 수 있었다. 또한 이들 무기화합물질의 상관성 분석결과 높은 상관성과 통계적인 유의성 (p<0.01)이 있었으며, 공통의 오염원을 추정 할 수 있었다. PM10의 가능한 오염원 기여도 평가에 있어서, 주거지역 A에서는 제철관련 오염원과 토양오염원의 공통 오염원 기여도가 33.4%로 나타나 산업관련 오염원의 기여도를 뚜렷하게 분리하여 평가하기는 곤란하였으나, 주거지역 B에서 토양관련 오염원의 독립적인 농도 기여도가 54%로 높게 나타났기 때문에, 상대적으로 산업지역과 인접한 주거지역 A에서 산업관련 오염원 기여도가 높은 것을 알 수 있었다.

• Composites Research
• /
• v.25 no.2
• /
• pp.34-39
• /
• 2012

A thermo-poro-elastic constitutive model of unidirectionally fiber-reinforced composite materials is suggested by extending the unmixing-mixing scheme which is based upon composite micromechanics. The strain components of thermal expansion due to a temperature change, gas pressure in pores, and chemical shrinkage are included in the constitutive model. On purpose to verify the derived constitutive relations, the representative volume element of two-dimensional lamina subject to various loading conditions is analyzed by the finite element method. The overall stress and strain responses are obtained, and compared with the predicted values by the unmixing-mixing scheme. The numerical results show the usefulness of the proposed model to predict the thermoelastic behavior of porous composites.

• Membrane Journal
• /
• v.12 no.3
• /
• pp.158-164
• /
• 2002

This study was carried out the fundamental experiment to investigate the pervaporation process through NaA zeolite membrane for recycling the wasted isopropyl alcohol(IPA) in semiconductor cleaning processes. The NaA zeolite membrane used showed the excellent separation performance for full range of feed concentration and at high temperature operation. At 80 in operation temperature and 90 wt% IPA in feed concentration the separation performance was obtained about $1,500 g/m^2/hr$ in the permeation flux and more than 1,000 in the separation factor. In continuous operation of dehydration of IPA the average permeation flux was obtained about $1,000 g/m^2/hr$ at 80 and 90 wt% IPA feed concentration.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.13 no.6
• /
• pp.438-444
• /
• 2003

A method using beamforming algorithm has been developed to measure oblique incidence reflection coefficients of sound absorption materials. MUSIC(multiple signal classification) method detects the angles of incidence and reflection. By separating the incident and reflected waves using beamforming method, the reflection coefficient is calculated. Spatial smoothing technique Is also used to reduce the coherence between the incident and reflected waves. Numerical and experiment results are performed to verify the accuracy of proposed method.