• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.5
• /
• pp.495-507
• /
• 2019

Data sparsity is one of the main challenges for the recommender system. The recommender system contains massive data in which only a small part is the observed data and the others are missing data. Most studies assume that missing data is randomly missing from the dataset. Therefore, they only use observed data to train recommendation model, then recommend items to users. In actual case, however, missing data do not lost randomly. In our research, treat these missing data as negative examples of users' interest. Three sample methods are seamlessly integrated into SVD++ algorithm and then propose SVD++_W, SVD++_R and SVD++_KNN algorithm. Experimental results show that proposed sample methods effectively improve the precision in Top-N recommendation over the baseline algorithms. Among the three improved algorithms, SVD++_KNN has the best performance, which shows that the KNN sample method is a more effective way to extract the negative examples of the users' interest.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.17 no.3
• /
• pp.20-25
• /
• 1980

A singular value decomposition (SVD) pseudo-inversion method has been applied to the image reconstruction from projections. This approach is relatively unknown and differs from conventionally used reconstructioll methods such as the Foxier convolution and iterative techniques. In this paper, two SVD pseudo-inversion methods have been discussed for the search of optimum reconstruction and restoration, one using truncated inverse filtering, the other scalar Wiener filtering. These methods partly overcome the ill-conditioned nature of restoration problems by trading off between noise and signal quality. To test the SVD pseudo-inversion method, simulations were performed from projection data obtained from a phantom using truncated inversefiltering. The results are presented together with some limitations particular to the applications of the method to the general class of 3-D image reconstruction and restoration.

• Journal of Information Processing Systems
• /
• v.13 no.6
• /
• pp.1585-1601
• /
• 2017

With the rapid development of computer technologies, a number of image modification methods have emerged, which have great impacts on the security of image information. Therefore, it is necessary to protect the integrity and authenticity of digital images, and digital watermarking technique consequently becomes a research hotspot. An effort is made to survey and analyze advancements of image watermarking algorithms based on singular value decomposition (SVD) in recent years. In the first part, an overview of watermarking techniques is presented and then mathematical theory of SVD is given. Besides, SVD watermarking model, features, and evaluation indexes are demonstrated. Various SVD-based watermarking algorithms, as well as hybrid watermarking algorithms based on SVD and other transforms for copyright protection, tamper detection, location, and recovery are reviewed in the last part.

• Journal of the Korean Society of Radiology
• /
• v.15 no.5
• /
• pp.591-602
• /
• 2021

Acute ischemic stroke(AIS) should be diagnosed within a few hours of onset of cerebral infarction symptoms using diagnostic radiology. In this study, we evaluated the clinical usefulness of SVD and the Bayesian algorithm to measure the volume of cerebral infarction using computed tomography perfusion(CTP) imaging and magnetic resonance diffusion-weighted imaging(MR DWI). We retrospectively included 50 patients (male : female = 33 : 17) who visited the emergency department with symptoms of AIS from September 2017 to September 2020. The cerebral infarct volume measured by SVD and the Bayesian algorithm was analyzed using the Wilcoxon signed rank test and expressed as a median value and an interquartile range of 25 - 75 %. The core volume measured by SVD and the Bayesian algorithm using was CTP imaging was 18.07 (7.76 - 33.98) cc and 47.3 (23.76 - 79.11) cc, respectively, while the penumbra volume was 140.24 (117.8 - 176.89) cc and 105.05 (72.52 - 141.98) cc, respectively. The mismatch ratio was 7.56 % (4.36 - 15.26 %) and 2.08 % (1.68 - 2.77 %) for SVD and the Bayesian algorithm, respectively, and all the measured values had statistically significant differences (p < 0.05). Spearman's correlation analysis showed that the correlation coefficient of the cerebral infarct volume measured by the Bayesian algorithm using CTP imaging and MR DWI was higher than that of the cerebral infarct volume measured by SVD using CTP imaging and MR DWI (r = 0.915 vs. r = 0.763 ; p < 0.01). Furthermore, the results of the Bland Altman plot analysis demonstrated that the slope of the scatter plot of the cerebral infarct volume measured by the Bayesian algorithm using CTP imaging and MR DWI was more steady than that of the cerebral infarct volume measured by SVD using CTP imaging and MR DWI (y = -0.065 vs. y = -0.749), indicating that the Bayesian algorithm was more reliable than SVD. In conclusion, the Bayesian algorithm is more accurate than SVD in measuring cerebral infarct volume. Therefore, it can be useful in clinical utility.

• Journal of the Institute of Convergence Signal Processing
• /
• v.16 no.4
• /
• pp.122-128
• /
• 2015

As multimedia information is spread over cyber networks, problems such as protection of legal rights and original proof of an information owner raise recently. Various image transformations of DCT, DFT and DWT have been used to embed a watermark as a token of ownership. Recently, SVD being used in the field of numerical analysis is additionally applied to the watermarking methods. A watermarking method is proposed in this paper using Gabor cosine and sine transform as well as SVD for embedding and extraction of watermarks for digital images. After delivering attacks such as noise addition, space transformation, filtering and compression on watermarked images, watermark extraction algorithm is performed using the proposed GCST-SVD method. Normalized correlation values are calculated to measure the similarity between embedded watermark and extracted one as the index of watermark performance. Also visual inspection for the extracted watermark images has been done. Watermark images are inserted into the lowest vertical ac frequency band. From the experimental results, the proposed watermarking method using the singular vectors of SVD shows large correlation values of 0.9 or more and visual features of an embedded watermark for various attacks.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.2 s.308
• /
• pp.82-86
• /
• 2006

The design of a pattern recognition system generally involves the three aspects: preprocessing, feature extraction, and decision making. Among them, a feature extraction method determines an appropriate subspace of dimensionality in the original feature space of dimensionality so that it can reduce the complexity of the system and help to improve successful recognition rates. Linear transforms, such as principal component analysis, factor analysis, and linear discriminant analysis have been widely used in pattern recognition for feature extraction. This paper shows that singular value decomposition (SVD) can be applied usefully in feature extraction stage of pattern recognition. As an application, a remote sensing problem is applied to verify the usefulness of SVD. The experimental result indicates that the feature extraction using SVD can improve the recognition rate about 25% compared with that of PCA.

• Journal of the Institute of Convergence Signal Processing
• /
• v.14 no.3
• /
• pp.154-161
• /
• 2013

In this paper, Gabor cosine and sine transform considered as human visual filter is applied to watermarking methods for digital images. Four algorithms by using singular values or principal components of SVD in the frequency domain are proposed for watermark embedding and extraction. Two dimensional image is used as an embedded watermark. To measure the similarity between the embedded watermark image and the extracted one, a normalized correlation value is computed for the comparison of the four proposed methods with various attacks. Extracted watermark images are also provided for visual inspection. The proposed GCST-SVD method which embeds a watermark image into the lowest vertical or horizontal ac frequency band can provide useful watermarking algorithm with high correlation values and visual watermark features from experimental results for various attacks.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.12
• /
• pp.8455-8461
• /
• 2015

In this paper, we propose new search algorithms using SVD(Singular Value Decomposition) for fast search of Raman spectrum. In the proposed algorithms, small number of the eigen vectors obtained by SVD are chosen in accordance with their respective significance to achieve computation reduction. By introducing pilot test, we exclude large number of data from search and then, we apply partial distance search(PDS) for further computation reduction. We prepared 14,032 kinds of chemical Raman spectrum as the library for comparisons. Experiments were carried out with 7 methods, that is Full Search, PDS, 1DMPS modified MPS for applying to 1-dimensional space data with PDS(1DMPS+PDS), 1DMPS with PDS by using descending sorted variance of data(1DMPS Sort with Variance+PDS), 250-dimensional components of the SVD with PDS(250SVD+PDS) and proposed algorithms, PSP and PSSP. For exact comparison of computations, we compared the number of multiplications and additions required for each method. According to the experiments, PSSP algorithm shows 64.8% computation reduction when compared with 250SVD+PDS while PSP shows 157% computation reduction.

• Korean Journal of Community Nutrition
• /
• v.12 no.4
• /
• pp.396-404
• /
• 2007

Increased oxidative stress contributes to the progression of atherosclerosis. We measured serum antioxidant mineral concentrations, capacities of serum antioxidant enzymes and fasting lipid profile in 97 male patients with coronary artery disease (CAD) and 21 male controls. Nutrient intake was assessed by the semi-quantitative food frequency method. CAD patients were divided into single-vessel disease (SVD, n=66) and multi-vessel disease (MVD, n = 31) groups on the coronary angiography. The ratio of serum LDL- to HDL-cholesterol elevated with an increasing number of diseased vessels compared to the control (control < SVD < MVD, p < 0.05). Patients with SVD and MVD had higher levels of serum lipoprotein (a) than the control (p < 0.05). The mean intake of carbohydrate, protein and cholesterol was higher in MVD patients and the intakes of vitamins C and E were lower in MVD and SVD patients than in the control (p < 0.05). Serum copper (Cu) and zinc (Zn) levels were higher in MVD and SVD patients than in the control (Cu: control $75.8{\pm}5.07$, SVD $99.2{\pm}2.90$, MVD $100.1{\pm}2.32{\mu}g/dL$, p<0.01; Zn: $76.8{\pm}5.36$, $119.0{\pm}5.95$, $129.1{\pm}2.70{\mu}g/dL$, p < 0.01). And the ratio of Zn to Cu was higher in SVD and MVD patients than in the control (control $0.78{\pm}0.06$, SVD $0.88{\pm}0.05$, MVD $0.99{\pm}0.04$, P < 0.05). The activity of glutathione peroxidase (GSH-Px) was lower in MVD than in SVD and the control (control $35.13{\pm}1.34$, SVD $35.30{\pm}1.01$, MVD $31.00{\pm}1.04 U/mg$ protein, p < 0.05). The ratio of the activities of superoxide dismutase (SOD) to GSH-Px was higher in MVD than in control and SVD (p < 0.05). In groups with CAD, serum Cu and Zn concentrations and their ratio were changed compared to the control. GSH-Px activity was decreased and the ratio of SOD to GSH-Px was increased in the patients with MVD. The balances between the activities of SOD and GSH-Px should also be considered a risk factor in CAD patients.

• Journal of Korean Ophthalmic Optics Society
• /
• v.1 no.1
• /
• pp.29-39
• /
• 1996

We studied the method which would determine the appropriate additive damping factor for the damped least sequres(DLS) optimization. We calculated eigenvalues of the product of the Jacobian matrix of error function by using the singular value decomposition(SVD) method. While suitable damping factor was appiled to the additive DLS by using SVD and Gaussian elimination method, the convergence and stability of the optimization process were examined in a triplet-type camera lens-system where the condition number is well conditioned. We compared the convergence and stability of merit function when median, maximum and minimum of eigenvalues were used as a damping factor in the optimization process. When damping factor is median of eigenvalue, the convergence and stability of merit function are more excellent than in the case of two other eigenvalues. Thus, we adopt the median of eigenvalues as an appropriate damping factor. Next, by using SVD and Gaussian elimination method, we compound the convergence and stability of optimization process for triplet-type camera lens-system design. In these two method; triplet-type camera lens-system in which condition number is well conditioned, has little improvement with the combination of DLS and SVD.