• Communications for Statistical Applications and Methods
• /
• 제20권6호
• /
• pp.481-490
• /
• 2013

Classification is an important research area as data can be easily obtained even if the number of predictors becomes huge. The support vector machine(SVM) is widely used to classify a subject into a predetermined group because it gives sound theoretical background and better performance than other methods in many applications. The SVM can be viewed as a penalized method with the hinge loss function and penalty functions. Instead of $L_2$ penalty function Fan and Li (2001) proposed the smoothly clipped absolute deviation(SCAD) satisfying good statistical properties. Despite the ability of SVMs, they have drawbacks of non-robustness when there are outliers in the data. We develop a robust SVM method using a weight function with the SCAD penalty function based on the local quadratic approximation. We compare the performance of the proposed SVM with the SVM using the $L_1$ and $L_2$ penalty functions.

• 정보처리학회논문지D
• /
• 제13D권4호
• /
• pp.455-462
• /
• 2006

고차원 데이터 공간에서의 효과적인 검색을 위해 최근 VA-file[1], LPC-file[2] 등과 같이 벡터 근사에 기반을 둔 필터링 색인 방법들이 연구되었다. 필터링 색인 방법은 벡터를 근사한 작은 크기의 색인 정보를 사용하여 근사 거리를 계산하고, 이를 사용하여 질의 벡터와 유사하지 않은 대부분의 벡터들을 빠른 시간 안에 검색 대상에서 제외한다. 즉, 실제 벡터 대신 근사 벡터를 읽어 디스크 I/O 시간을 줄여 전체 검색 속도를 향상시키는 것이다. 하지만 VA-file 이나 LPC-file은 근사 거리를 구하는 방법이 순차 검색과 같거나 복잡하기 때문에 검색 속도 향상 효과가 그리 크지 않다는 문제점을 가지고 있다. 본 논문은 이러한 근사 거리 계산 시간을 줄이기 위하여 새로운 비트맵 색인 구조를 제안한다. 근사 거리 계산속도의 향상을 위하여, 각 객체의 값을 특성 벡터 공간상의 위치를 나타내는 비트 패턴으로 저장하고, 객체 사이의 거리를 구하는 연산은 실제 벡터 값의 연산보다 속도가 훨씬 빠른 XOR 비트 연산으로 대체한다. 실험에 의하면 본 논문이 제안하는 방법은 기존 벡터 근사 접근 방법들과 비교하여 데이터 읽기시간은 더 크지만, 계산 시간을 크게 줄임으로써 전체 검색 속도는 순차 검색의 약 4배, 기존의 방법들보다는 최대 2배의 성능이 향상되었다. 결과적으로, 데이터베이스의 속도가 충분히 빠른 경우 기존의 벡터 근사 접근법의 필터링을 위한 계산 시간을 줄임으로써 더욱 검색 성능을 향상 시킬 수 있음을 확인할 수 있다.

• Spatial Information Research
• /
• 제13권4호
• /
• pp.355-364
• /
• 2005

모바일 기기의 제한적 환경에서 공간데이터의 활용을 극대화하기 위해 벡터데이터의 압축에 대한 연구가 최근 이뤄지고 있다. 이 중 군집화 방법을 이용한 벡터데이터 압축은 기존 압축방법과 다른 새로운 형태로 주목을 받고 있다. 그러나 현재까지 연구는 데이터의 동적인 갱신이 고려되지 않았다. 본 연구는 기존의 군집화 방법을 이용한 벡터데이터 압축방법의 문제점을 파악하고, 데이터의 동적인 갱신이 고려된 압축 방법을 제시하였다. 실험을 통한 결과는 갱신이 발생하였을 경우 제안된 방법이 더 좋은 결과를 나타냄을 확인할 수 있었다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2009년도 정보 및 제어 심포지움 논문집
• /
• pp.320-322
• /
• 2009

Support vector learning attracts great interests in the areas of pattern classification, function approximation, and abnormality detection. In this pater, we design the controller using support vector regression which has good properties in comparison with multi-layer perceptron or radial basis function. The applicability of the presented method is illustrated via an example simulation.

• International Journal of CAD/CAM
• /
• 제13권1호
• /
• pp.1-11
• /
• 2013

Just by adjusting the control points iteratively, progressive iterative approximation (PIA) presents an intuitive and straightforward scheme such that the resulting limit curve (surface) can interpolate the original data points. In order to obtain more flexibility, adjusting only a subset of the control points, a new method called local progressive iterative approximation (LPIA) has also been proposed. But to this day, there are two problems about PIA and LPIA: (1) Only an approximation process is discussed, but the accurate convergence curves (surfaces) are not given. (2) In order to obtain an interpolating curve (surface) with high accuracy, recursion computations are needed time after time, which result in a large workload. To overcome these limitations, this paper gives an explicit matrix expression of the control points of the limit curve (surface) by the PIA or LPIA method, and proves that the column vector consisting of the control points of the PIA's limit curve (or surface) can be obtained by multiplying the column vector consisting of the original data points on the left by the inverse matrix of the collocation matrix (or the Kronecker product of the collocation matrices in two direction) of the blending basis at the parametric values chosen by the original data points. Analogously, the control points of the LPIA's limit curve (or surface) can also be calculated by one-step. Furthermore, the $G^1$ joining conditions between two adjacent limit curves obtained from two neighboring data points sets are derived. Finally, a simple LPIA method is given to make the given tangential conditions at the endpoints can be satisfied by the limit curve.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2004년도 추계학술대회 학술발표 논문집 제14권 제2호
• /
• pp.179-191
• /
• 2004

• 한국수학교육학회지시리즈A:수학교육
• /
• 제22권2호
• /
• pp.41-44
• /
• 1984

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2015년도 학술발표회
• /
• pp.475-478
• /
• 2015

The objective of this study is to apply a hybrid model for estimating solar radiation and investigate their accuracy. A hybrid model is wavelet-based support vector machines (WSVMs). Wavelet decomposition is employed to decompose the solar radiation time series into approximation and detail components. These decomposed time series are then used as inputs of support vector machines (SVMs) modules in the WSVMs model. Results obtained indicate that WSVMs can successfully be used for the estimation of daily global solar radiation at Champaign and Springfield stations in Illinois.

• 한국경영과학회지
• /
• 제18권3호
• /
• pp.13-34
• /
• 1993

By operating on many part of a software system concurrently, the parallel processing computers may provide several orders of magnitude more computing power than traditional serial computers. If the Lagrangean approximation procedure is applied to a large scale manufacturing problem which is decomposable into many subproblems, the procedure is a perfect candidate for parallel processing. By distributing Lagrangean subproblems for given multiplier to multiple processors, concurrently running processors and modifying Lagrangean multipliers at the end of each iteration of a subgradient method,a parallel processing of a Lagrangean approximation procedure may provide a significant speedup. This purpose of this research is to investigate the potential of the parallelized Lagrangean approximation procedure (PLAP) for certain combinational optimization problems in manufacturing systems. The framework of a Plap is proposed for some combinatorial manufacturing problems which are decomposable into well-structured subproblems. The synchronous PLAP for the multistage dynamic lot-sizing problem is implemented on a parallel computer Alliant FX/4 and its computational experience is reported as a promising application of vector-concurrent computing.

• 대한기계학회논문집
• /
• 제19권2호
• /
• pp.433-442
• /
• 1995

For the line search of a multi-variable optimization, an efficient algorithm is presented. The algorithm sequentially employs several polynomial approximations such as 2-point quadratic interpolation, 3-point cubic interpolation/extrapolation and 4-point cubic interpolation/extrapolation. The order of polynomial function is automatically increased for improving the accuracy of approximation. The method of approximation (interpolation or extrapolation) is automatically switched by checking the slope information of the sample points. Also, for selecting the initial step length along the descent vector, a new approach is presented. The performance of the proposed method is examined by solving typical test problems such as mathematical problems, mechanical design problems and dynamic response problems.