• 유통과학연구
• /
• 제14권9호
• /
• pp.25-29
• /
• 2016

Purpose - Finding top K persuaders in consumer network is an important problem in marketing. Recently, a new method of computing persuasion scores, interpreted as fixed point or stable distribution for given persuasion probabilities, was proposed. Top K persuaders are chosen according to the computed scores. This research proposed a new definition of persuasion scores relaxing some conditions on the matrix of probabilities, and a method to identify top K persuaders based on the defined scores. Research design, data, and methodology - A new method of computing top K persuaders is computed by singular value decomposition (SVD) of the matrix which represents persuasion probabilities between entities. Results - By testing a randomly generated instance, it turns out that the proposed method is essentially different from the previous study sharing a similar idea. Conclusions - The proposed method is shown to be valid with respect to both theoretical analysis and empirical test. However, this method is limited to the category of persuasion scores relying on the matrix-form of persuasion probabilities. In addition, the strength of the method should be evaluated via additional experiments, e.g., using real instances, different benchmark methods, efficient numerical methods for SVD, and other decomposition methods such as NMF.

• Structural Engineering and Mechanics
• /
• 제14권2호
• /
• pp.151-170
• /
• 2002

Moving force identification is a very important inverse problem in structural dynamics. Most of the identification methods are eventually converted to a linear algebraic equation set. Different ways to solve the equation set may lead to solutions with completely different levels of accuracy. Based on the measured bending moment responses of the bridge made in laboratory, this paper presented the time domain method (TDM) and frequency-time domain method (FTDM) for identifying the two moving wheel loads of a vehicle moving across a bridge. Directly calculating pseudo-inverse (PI) matrix and using the singular value decomposition (SVD) technique are adopted as means for solving the over-determined system equation in the TDM and FTDM. The effects of bridge and vehicle parameters on the TDM and FTDM are also investigated. Assessment results show that the SVD technique can effectively improve identification accuracy when using the TDM and FTDM, particularly in the case of the FTDM. This improved accuracy makes the TDM and FTDM more feasible and acceptable as methods for moving force identification.

• Structural Engineering and Mechanics
• /
• 제25권4호
• /
• pp.383-403
• /
• 2007

The contact problem for an elastic layer resting on an elastic half plane is considered according to the theory of elasticity with integral transformation technique. External loads P and Q are transmitted to the layer by means of two dissimilar rigid flat punches. Widths of punches are different and the thickness of the layer is h. All surfaces are frictionless and it is assumed that the layer is subjected to uniform vertical body force due to effect of gravity. The contact along the interface between elastic layer and half plane will be continuous, if the value of load factor, ${\lambda}$, is less than a critical value, ${\lambda}_{cr}$. However, if tensile tractions are not allowed on the interface, for ${\lambda}$ > ${\lambda}_{cr}$ the layer separates from the interface along a certain finite region. First the continuous contact problem is reduced to singular integral equations and solved numerically using appropriate Gauss-Chebyshev integration formulas. Initial separation loads, ${\lambda}_{cr}$, initial separation points, $x_{cr}$, are determined. Also the required distance between the punches to avoid any separation between the punches and the layer is studied and the limit distance between punches that ends interaction of punches, is investigated. Then discontinuous contact problem is formulated in terms of singular integral equations. The numerical results for initial and end points of the separation region, displacements of the region and the contact stress distribution along the interface between elastic layer and half plane is determined for various dimensionless quantities.

• 한국광학회지
• /
• 제6권3호
• /
• pp.178-187
• /
• 1995

렌즈 설계에 적합한 감쇠계수를 결정하는 방법을 연구하였다. 구해진 감쇠계수를 additive 감쇠최소 자승법에 도입하였을 때, triplet형 사진렌즈 설계에서 최적화 과정의 수렴성과 안정성에 관해서 조사하였다. 오차함수의 Jacobian 행열의 곱에 대한 고유값을 SVD(singular value decomposition)를 통해 구한 후 고유값들의 중간치를 적합한 감쇠계수로 결정하였다. 적합한 감쇠계수를 이용하여 triplet형 사진렌즈를 최적화한 결과 수렴성과 안정성이 향상되었다. 직교변환 방법으로 Jacobian 행열의 해를 구하면 정규 방정식을 사용하여 해를 구할 때 발생하는 수치적 부정확성을 개선할 수 있었다. 적합한 감쇠계수 선택법과 Jacobian 행열의 직교변환 방법을 같이 사용하는 것이 고차의 항들을 가지는 비구면 렌즈 설계에 유용함을 알 수 있었다.

• 한국산업정보학회논문지
• /
• 제13권3호
• /
• pp.1-8
• /
• 2008

교통 사고, 화재, 범죄 등이 발생했을 때, 누구든지 사진을 찍을 수 있고, 그 사진이 증거로 사용될 수 있다. 디지털 워터마킹은 특정 카메라 폰에서 찍은 사진이 인증될 수 있도록 보증할 수 있게 해준다. 본 논문에서는 카메라 폰에 적합한 디지털 워터마킹 방법을 제안한다. 이 방법은 특정치 분해와 웨이블릿 분해를 기반으로 한다. 실험 결과는 제안된 방법이 JPEG 압축에 대해 보안성과 강인성이 좋음을 보여준다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권5호
• /
• pp.1708-1727
• /
• 2021

In the advent of the twenty-first century, human beings began to closely interact with technology. Today, technology is developing, and as a result, the world wide web (www) has a very important place on the Internet and the significant task is fulfilled by Web services. A lot of Web services are available on the Internet and, therefore, it is difficult to find matching Web services among the available Web services. The recommendation systems can help in fixing this problem. In this paper, our observation was based on the recommended method such as the collaborative filtering (CF) technique which faces some failure from the data sparsity and the cold-start problems. To overcome these problems, we first applied an ontology-based clustering and then the k-nearest neighbor (KNN) algorithm for each separate cluster group that effectively increased the data density using the past user interests. Then, user ratings were predicted based on the model-based approach, such as singular value decomposition (SVD) and the predictions used for the recommendation. The evaluation results showed that our proposed approach has a less prediction error rate with high accuracy after analyzing the existing recommendation methods.

• 한국산업정보학회논문지
• /
• 제15권2호
• /
• pp.49-57
• /
• 2010

본 논문에서는 역전파 신경망 알고리즘(BPNN: Back Propagation Neural Network)과 Singular Value Decomposition(SVD)를 이용하는 한글 문서 분류 시스템을 제안한다. BPNN은 학습을 통하여 만들어진 네트워크를 이용하여 문서분류를 수행한다. 이 방법의 어려움은 분류기에 입력되는 특징 공간이 너무 크다는 것이다. SVD를 이용하면 고차원의 벡터를 저차원으로 줄일 수 있고, 또한 의미있는 벡터 공간을 만들어 단어 사이의 중요한 관계성을 구축할 수 있다. 본 논문에서 제안한 BPNN의 성능 평가를 위하여 한국일보-2000/한국일보-40075 문서범주화 실험문서집합의 데이터 셋을 이용하였다. 실험결과를 통하여 BPNN과 SVD를 사용한 시스템이 한글 문서 분류에 탁월한 성능을 가지는 것을 보여준다.

• 인터넷정보학회논문지
• /
• 제25권1호
• /
• pp.49-56
• /
• 2024

There have been emerging many use-cases applying recommendation systems especially in online platform. Although the performance of recommendation systems is affected by a variety of factors, selecting appropriate features is difficult since most of recommendation systems have sparse data. Conventional matrix factorization (MF) method is a basic way to handle with problems in the recommendation systems. However, the MF based scheme cannot reflect non-linearity characteristics well. As deep learning technology has been attracted widely, a deep neural network (DNN) framework based collaborative filtering (CF) was introduced to complement the non-linearity issue. However, there is still a problem related to feature embedding for use as input to the DNN. In this paper, we propose an effective method using singular value decomposition (SVD) based feature embedding for improving the DNN performance of recommendation algorithms. We evaluate the performance of recommendation systems using MovieLens dataset and show the proposed scheme outperforms the existing methods. Moreover, we analyze the performance according to the number of latent features in the proposed algorithm. We expect that the proposed scheme can be applied to the generalized recommendation systems.

• International Journal of Railway
• /
• 제2권3호
• /
• pp.118-123
• /
• 2009

Railway track alignment Standards set a minimum lenght value for straight and circular alignments (art. 5.2.9.), in order to ensure passenger ride comfort in railway vehicles of which dynamic oscillations will thus have to be limited. The transitions between alignments can cause abrupt changes (usually called discontinuities or singular points of the alignment) of curvature, of rate of change of curvature or of rate of change of cant. A passenger is likely to experience effects due to the excitation of the elastic suspension of the vehicle which generates oscillations that are damped as the vehicle moves away from the singularity. The amplitude of these oscillations should be adequately attenuated by the damping of the suspension system within the interval between two successive singular points, especially to avoid resonances. Therefore minimum lengths between two successive singular points are stated in alignment standards. Nevertheless, these nonnative values can be overly conservative in some cases. As an alternative, track alignment designers could try to assess how much the excitation has been attenuated between two successive singular points and thus assess at which point a new singularity may be present without affecting ride comfort. Although such assessment can be made with commercial SW packages which simulate the dynamic behavior of a vehicle considered as a set of rigid bodies interconnected with elastic elements simulating the suspension systems (such as SIMPACK, ADAMS or VAMPIRE), a simplified and user-friendly computation method (based upon the analytical solution of differential equations governing the phenomenon) is made available in this paper to track design engineers, not always used to working with full dynamic models.

• 융합신호처리학회논문지
• /
• 제16권4호
• /
• pp.122-128
• /
• 2015

멀티미디어 정보들이 인터넷 공간에 확산됨에 따라서 원래 정보 소유자의 권리 보호와 원본 증명 등의 문제가 대두되고 있다. DCT, DFT, DWT 등의 여러 영상 변환들을 이용하여 소유권의 징표로 워터마크를 원본 영상에 삽입하는 방법을 많이 사용하였으나, 보다 최근에는 수치해석 분야에 많이 쓰이는 SVD(Singular Value Decomposition) 방법을 부가적으로 사용하고 있다. 본 논문에서는 SVD의 특이 벡터와 동시에 Gabor 코사인과 사인 변환을 이용하여 디지털 표지 영상에 워터마크를 삽입하고 추출하는 방법을 제안한다. 워터마크가 삽입된 영상에 잡음, 공간 변형, 필터링, 압축 등의 공격을 가한 후, GCST-SVD의 워터마크 추출 알고리즘을 적용한다. 워터마킹 성능을 평가하기 위해서 삽입한 워터마크와 추출한 워터마크 사이의 유사성 척도로써 정규화한 상관계수값을 측정한다. 또한 추출한 워터마크 영상으로부터 시각적으로 직접 원본 워터마크인지를 판단한다. 가장 낮은 수직 교류 주파수 대역에 워터마크를 삽입한 실험으로부터 SVD의 특이 벡터를 이용한 워터마킹 방법은 대부분 공격에서 0.9이상의 큰 상관값과 삽입한 워터마크의 특징들을 시각적으로 파악할 수 있었다.