• ETRI Journal
• /
• v.37 no.6
• /
• pp.1251-1258
• /
• 2015

We investigate an image recovery method for sparse-view computed tomography (CT) using an iterative shrinkage algorithm based on a second-order approach. The two-step iterative shrinkage-thresholding (TwIST) algorithm including a total variation regularization technique is elucidated to be more robust than other first-order methods; it enables a perfect restoration of an original image even if given only a few projection views of a parallel-beam geometry. We find that the incoherency of a projection system matrix in CT geometry sufficiently satisfies the exact reconstruction principle even when the matrix itself has a large condition number. Image reconstruction from fan-beam CT can be well carried out, but the retrieval performance is very low when compared to a parallel-beam geometry. This is considered to be due to the matrix complexity of the projection geometry. We also evaluate the image retrieval performance of the TwIST algorithm -sing measured projection data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.10
• /
• pp.5015-5038
• /
• 2018

In this paper, a novel spectral-spatial joint sparse representation algorithm for hyperspectral image classification is proposed based on multi-layer superpixels in various scales. Superpixels of various scales can provide complete yet redundant correlated information of the class attribute for test pixels. Therefore, we design a joint sparse model for a test pixel by sampling similar pixels from its corresponding superpixels combinations. Firstly, multi-layer superpixels are extracted on the false color image of the HSI data by principal components analysis model. Secondly, a group of discriminative sampling pixels are exploited as reconstruction matrix of test pixel which can be jointly represented by the structured dictionary and recovered sparse coefficients. Thirdly, the orthogonal matching pursuit strategy is employed for estimating sparse vector for the test pixel. In each iteration, the approximation can be computed from the dictionary and corresponding sparse vector. Finally, the class label of test pixel can be directly determined with minimum reconstruction error between the reconstruction matrix and its approximation. The advantages of this algorithm lie in the development of complete neighborhood and homogeneous pixels to share a common sparsity pattern, and it is able to achieve more flexible joint sparse coding of spectral-spatial information. Experimental results on three real hyperspectral datasets show that the proposed joint sparse model can achieve better performance than a series of excellent sparse classification methods and superpixels-based classification methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.5
• /
• pp.1163-1185
• /
• 2024

With the sharp increase in the volume of literature data, researchers must spend considerable time and energy locating desired papers. A paper recommendation is the means necessary to solve this problem. Unfortunately, the large amount of data combined with sparsity makes personalizing papers challenging. Traditional matrix decomposition models have cold-start issues. Most overlook the importance of information and fail to consider the introduction of noise when using side information, resulting in unsatisfactory recommendations. This study proposes a paper recommendation method (PR-SLSMF) using document-level representation learning with citation-informed transformers (SPECTER) and low-rank and sparse matrix factorization; it uses SPECTER to learn paper content representation. The model calculates the similarity between papers and constructs a weighted heterogeneous information network (HIN), including citation and content similarity information. This method combines the LSMF method with HIN, effectively alleviating data sparsity and cold-start issues and avoiding topic drift. We validated the effectiveness of this method on two real datasets and the necessity of adding side information.

• KIISE Transactions on Computing Practices
• /
• v.20 no.11
• /
• pp.610-614
• /
• 2014

Like many types of scientific data, results from simulations of volcanic ash diffusion are of a clustered sparse matrix in the netCDF format. Since these data sets are large in size, they generate high storage and transmission costs. In this paper, we suggest a new method that reduces the size of the data of volcanic ash diffusion simulations by converting the multi-dimensional index to a single dimension and keeping only the starting point and length of the consecutive zeros. This method presents performance that is almost as good as that of ZIP format compression, but does not destroy the netCDF structure. The suggested method is expected to allow for storage space to be efficiently used by reducing both the data size and the network transmission time.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.2
• /
• pp.205-211
• /
• 2012

The retrieved documents have to be transformed into proper data structure for the clustering algorithms of statistics and machine learning. A popular data structure for document clustering is document-term matrix. This matrix has the occurred frequency value of a term in each document. There is a sparsity problem in this matrix because most frequencies of the matrix are 0 values. This problem affects the clustering performance. The sparseness of document-term matrix decreases the performance of clustering result. So, this research uses the factor score by factor analysis to solve the sparsity problem in document clustering. The document-term matrix is transformed to document-factor score matrix using factor scores in this paper. Also, the document-factor score matrix is used as input data for document clustering. To compare the clustering performances between document-term matrix and document-factor score matrix, this research applies two typed matrices to self organizing map (SOM) clustering.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1991.10a
• /
• pp.23-23
• /
• 1991

• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.329-345
• /
• 2021

Recommender systems use data from customers to suggest personalized products. The recommender systems can be categorized into three cases; collaborative filtering, contents-based filtering, and hybrid recommender system that combines the first two filtering methods. In this work, we introduce and compare deep learning-based recommender system using autoencoder. Autoencoder is an unsupervised deep learning that can effective solve the problem of sparsity in the data matrix. Five versions of autoencoder-based deep learning models are compared via three real data sets. The first three methods are collaborative filtering and the others are hybrid methods. The data sets are composed of customers' ratings having integer values from one to five. The three data sets are sparse data matrix with many zeroes due to non-responses.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.1
• /
• pp.1-9
• /
• 2021

OpenGL compute shader is a shader stage that operate differently from other shader stage and it can be used for the calculating purpose of any data in parallel. This paper proposes a GPU-based parallel algorithm for computing sparse linear systems through conjugate gradient using an iterative method, which perform calculation on OpenGL compute shader. Basically, this sparse linear solver is used to solve large linear systems such as symmetric positive definite matrix. Four well-known matrix formats (Dense, COO, ELL and CSR) have been used for matrix storage. The performance comparison from our experimental tests using eight sparse matrices shows that GPU-based linear solving system much faster than CPU-based linear solving system with the best average computing time 0.64ms in GPU-based and 15.37ms in CPU-based.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.32A no.9
• /
• pp.1244-1249
• /
• 1995

Wavelet transform technique is applied to two important electromagnetic problems:1) to analyze the frequency-domain radar echo from finite-size targets and 2) to the integral solution of two- dimensional electromagnetic scattering problems. Since the frequency- domain radar echo consists of both small-scale natural resonances and large-scale scattering center information, the multiresolution property of the wavelet transform is well suited for analyzing such ulti-scale signals. Wavelet analysis examples of backscattered data from an open- ended waveguide cavity are presented. The different scattering mechanisms are clearly resolved in the wavelet-domain representation. In the wavelet transform domain, the moment method impedance matrix becomes sparse and sparse matrix algorithms can be utilized to solve the resulting matrix equationl. Using the fast wavelet transform in conjunction with the conjugate gradient method, we present the time performance for the solution of a dihedral corner reflector. The total computational time is found to be reduced.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.7
• /
• pp.956-962
• /
• 2020

Recently, as the size of data used in an embedded system increases, the need for an ECC decoding operation to robustly receive a massive data is emphasized. In this paper, we propose a method to accelerate the execution of computations that derive syndrome vectors when ECC decoding is performed using Hamming code in an embedded system with a built-in GPU. The proposed acceleration method uses the matrix-vector multiplication of the decoding operation using the CSR format, one of the data structures representing sparse matrix, and is performed in parallel in the CUDA kernel of the GPU. We evaluated the proposed method using a target embedded board with a GPU, and the result shows that the execution time is reduced when ECC decoding operation accelerated based on the GPU than used only CPU.