• 한국IT서비스학회지
• /
• 제14권4호
• /
• pp.121-135
• /
• 2015

Ensemble classification is to utilize multiple classifiers instead of using a single classifier. Recently ensemble classifiers have attracted much attention in data mining community. Ensemble learning techniques has been proved to be very useful for improving the prediction accuracy. Bagging, boosting and random subspace are the most popular ensemble methods. In random subspace, each base classifier is trained on a randomly chosen feature subspace of the original feature space. The outputs of different base classifiers are aggregated together usually by a simple majority vote. In this study, we applied the random subspace method to the bankruptcy problem. Moreover, we proposed a method for optimizing the random subspace ensemble. The genetic algorithm was used to optimize classifier subset of random subspace ensemble for bankruptcy prediction. This paper applied the proposed genetic algorithm based random subspace ensemble model to the bankruptcy prediction problem using a real data set and compared it with other models. Experimental results showed the proposed model outperformed the other models.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2002년도 ICCAS
• /
• pp.75.1-75
• /
• 2002

Several model order reduction methods for large RLC circuits have been developed in the last few years. Krylop subspace based methods are extremely effective for generating the low order models of large system but there is no optimal theory for the resulting models. Alternatively, methods based truncated balanced realization have an optimality property but are too computationally expensive to use on complicated problems such as large RLC circuits. In this paper, we present a method for improving time domain response of reduced order RLC circuits. The method used here is based on combing Krylop subspace based method and truncated balanced realization method plus residualization. The metho...

• 한국항행학회논문지
• /
• 제26권3호
• /
• pp.166-171
• /
• 2022

본 논문에서는 선형 배열 균일 안테나에 입사하는 신호들의 상관행렬을 강건하게 생성하여 부공간 기반 기법의 도래각 추정 성능을 향상시키는 알고리즘을 제안한다. 기존의 부공간 기반 도래각 추정 기법은 상관행렬을 구한 후 신호 부공간과 잡음 부공간으로 분리하여 도래각을 추정한다. 그러나, 낮은 SNR, 작은 개수의 스냅샷에서 구해지는 상관행렬의 성분은 안테나의 잡음 성분으로 인하여 신호 부공간을 부정확하게 추정하여 도래각 추정 성능을 저하시킨다. 따라서, 기존의 상관행렬로부터 구해지는 가상의 신호 벡터를 슬라이딩 방식으로 배열함으로써 강건한 상관행렬을 생성한다. 기존의 상관행렬과 제안하는 강건한 상관행렬의 비교 분석을 위하여, 부공간 기반 기법의 대표적 방법인 MUSIC, ESPRIT을 이용하였다. 시뮬레이션 결과, 계산 복잡도는 기존의 상관행렬 대비 2.5% 이내 증가하였으나, 도래각 추정성능은 RMSE 1° 기준 SNR이 MUSIC, ESPRIT 모두 3dB 이상의 우수한 도래각 추정 성능을 보였다.

• 대한전자공학회논문지SP
• /
• 제47권6호
• /
• pp.162-170
• /
• 2010

원하는 신호의 도래방향에 관한 정보를 이용하여 적응 어레이는 이 방향으로 빔 이득을 유지하면서 간섭신호를 제거한다. 신호 부공간에서 가중벡터를 조정하면 전체 공간에서 조정하는 방식에 비해 빠른 수렴속도를 가지며, 도래각 정보에서의 에러에 강인한 특성을 가진다. 그러나 공분산 행렬의 고유분해가 필요하고 이에 따른 계산이 복잡하다. 본 논문에서는 PASTd(projection approximation subspace tracking with deflation) 방식에 기초하여 계산이 간단한 신호 부공간에 기초한 적응어레이를 제시한다. 제시된 방식은 고유벡터가 직교하도록 원래의 PASTd를 변형해서 사용하고 있고, 직접 고유분해하는 방식과 동일한 성능을 가지면서 계산량을 크게 감소시킬 수 있다. 또한 신호 부공간 어레이의 SINR(signal-to-interference plus noise ratio)성능을 이론적으로 분석하여, 이의 동작특성을 고찰하였다.

• 한국음향학회지
• /
• 제27권7호
• /
• pp.372-378
• /
• 2008

음성 검출기 (VAD, Voice Activity Detector)는 이동 통신이나 음성신호처리 등에 매우 중요한 기법으로 사용된다. 일반적인 음성 검출방식은 이산 푸리에 변환 (DFT, Discrete Fourier Transform)영역에서 통계적인 모델을 기반으로 하여 우도비검정 (LRT, Likelihood Ratio Test)을 하게 된다. 그리고 이 값을 임계값과 비교하며 음성인지 아닌지 판단하게 된다. 본 논문에서는 신호 준공간 (Signal Subspace)에 기반한 새로운 통계적 음성 검출 기법을 제안하다. 확률적인 주성분 분석 (PPCA, Probabilistic Principal Component Analysis)은 신호 준공간 방법에서 잡음신호에 대한 확률적인 모델을 얻기 위해 사용된다. 제안된 기법은 신호 준공간 영역에서 우도비검정에 기반을 두는 결정규칙을 적용하였다. 음성 검출 실험 결과는 신호 준공간 모델에 근거한 음성 검출기 기법이 주파수 영역에 기반한 가우시안 (Gaussian) 음성 검출기 보다 향상된 검출 결과를 보여준다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2000년도 제15차 학술회의논문집
• /
• pp.525-525
• /
• 2000

In this paper, we present recursive algorithms for state space model identification using subspace extraction via Schur complement. It is shown that an estimate of the extended observability matrix can be obtained by subspace extraction via Schur complement. A relationship between the least squares residual and the Schur complement matrix obtained from input-output data is shown, and the recursive algorithms for the subspace-based state-space model identification (4SID) methods are developed. We also proposed the above algorithm for an instrumental variable (IV) based 4SID method. Finally, a numerical example of the application of the algorithms is illustrated.

• Structural Monitoring and Maintenance
• /
• 제4권4호
• /
• pp.365-379
• /
• 2017

This paper presents a comparison of the black-box and the physics based derived gray-box models for subspace identification for structures subjected to support-excitation. The study compares the damage detection capabilities of both these methods for linear time invariant (LTI) systems as well as linear time-varying (LTV) systems by extending the gray-box model for time-varying systems using short-time windows. The numerically simulated IASC-ASCE Phase-I benchmark building has been used to compare the two methods for different damage scenarios. The efficacy of the two methods for the identification of stiffness parameters has been studied in the presence of different levels of sensor noise to simulate on-field conditions. The proposed extension of the gray-box model for LTV systems has been shown to outperform the black-box model in capturing the variation in stiffness parameters for the benchmark building.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권4호
• /
• pp.703-719
• /
• 2011

In multi-user wireless communication systems, adaptive modulation and scheduling are promising techniques for increasing the system throughput. However, a mass of wireless recourse will be occupied and spectrum efficiency will be decreased to feedback channel quality indication (CQI) of all users in every subcarrier or chunk for adaptive orthogonal frequency division multiplexing (OFDM) systems. Thus numerous limited feedback schemes are proposed to reduce the system overhead. The recently proposed compressive sensing (CS) theory provides a new framework to jointly measure and compress signals that allows less sampling and storage resources than traditional approaches based on Nyquist sampling. In this paper, we proposed two novel CQI feedback schemes based on general CS and subspace CS, respectively, both of which could be used in a wireless OFDM system. The feedback rate with subspace CS is greatly decreased by exploiting the subspace information of the underlying signal. Simulation results show the effectiveness of the proposed methods, with the same feedback rate, the throughputs with subspace CS outperform the discrete cosine transform (DCT) based method which is usually employed, and the throughputs with general CS outperform DCT when the feedback rate is larger than 0.13 bits/subcarrier.

• 한국자동차공학회논문집
• /
• 제11권5호
• /
• pp.210-218
• /
• 2003

MDO (multidisciplinary design optimization) technology has been proposed and applied to solve large and complex optimization problems where multiple disciplinaries are involved. In this research. an MDO problem is defined for automobile design which has crashworthiness analyses. Crash model which are consisted of airbag, belt integrated seat (BIS), energy absorbing steering system .and safety belt is selected as a practical example for MDO application to vehicle system. Through disciplinary analysis, vehicle system is decomposed into structure subspace and occupant subspace, and coupling variables are identified. Before subspace optimization, values of coupling variables at given design point must be determined with system analysis. The system analysis in MDO is very important in that the coupling between disciplines can be temporary disconnected through the system analysis. As a result of system analysis, subspace optimizations are independently conducted. However, in vehicle crash, system analysis methods such as Newton method and fixed-point iteration can not be applied to one. Therefore, new system analysis algorithm is developed to apply to crashworthiness. It is conducted for system analysis to determine values of coupling variables. MDO algorithm which is applied to vehicle crash is MDOIS (Multidisciplinary Design Optimization Based on Independent Subspaces). Then, structure and occupant subspaces are independently optimized by using MDOIS.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 하계학술대회 논문집 D
• /
• pp.2143-2145
• /
• 2003

In the linear system identification using the discrete time constant coefficients, there is a subspace method based on 4SID recently much suggested instead of the parametric method like as the maximum likelihood method. The subspace method is not related with the impulse response and difference equation in its input-output equation, but with the system matrix of the direct state space model from the input-output data. The subspace method is a very useful tool to adopt in the multivariable system identification, but it has a shortage unable to adopt in the closed-loop system identification. In this paper, we are suggested the methods to get rid of the shortage of the subspace method in the closed-loop system identification. The subspace method is used in the estimate of the output prediction values from the estimating of the state space vector. And we have compared the results with the outputs of the recursive least square method in the numerical simulation.