• 대한전자공학회논문지SP
• /
• 제46권5호
• /
• pp.15-24
• /
• 2009

Compressed sensing은 소수의 선형 관측으로부터 sparse 신호를 복원하는 문제를 언급하고 있다. 최근 벡터 경우에서의 성공적인 연구 결과가 행렬의 경우로 확장되었다. Low-rank 행렬의 compressed sensing은 ill-posed inverse problem을 low-rank 정보를 이용하여 해결한다. 본 문제는 rank 최소화 혹은 low-rank 근사의 형태로 나타내질 수 있다. 본 논문에서는 최근 제안된 여러 가지 효율적인 알고리즘에 대한 survey를 제공한다.

• 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.

• 한국전산구조공학회논문집
• /
• 제23권5호
• /
• pp.475-482
• /
• 2010

본 논문에서는 대규모 자유도 시스템의 병렬처리를 위하여 2단계로 이루어진 영역분할법(Domain Decomposition Method) 기반의 병렬 알고리즘을 제안하였다. 분할된 영역의 내부 및 외부 경계를 상위영역문제로 정의하고 국부영역문제는 변위 경계조건이 모두 주어지는 분할영역에서의 Dirichlet 문제로 구성한다. 상위영역에서는 전체 상위영역에 대한 강성 행렬의 어셈블이 필요없는 반복법을 통하여 변위를 구하고, 이를 바탕으로 국부영역에서 Multi-Frontal Sparse Solver (MFSS)를 이용하여 변위를 계산한다. 상위영역문제의 연산에서 프로세서 간의 데이터 교환을 최소화하여 계산효율을 유지하며, 동시에 해석 가능한 자유도를 증대시키는 병렬 PCG(Preconditioned Conjugate Gradient)법 기반의 알고리즘을 개발하였다. 제안된 알고리즘을 적용하여 수치해석을 수행한 결과, 프로세서 수가 증가할수록 계산성능의 손실없이 해석 가능한 자유도가 비례하여 증가하는 선형 확장성을 관찰할 수 있었으며, 대규모 자유도 문제에 효과적으로 사용 가능함을 확인하였다.

• Current Optics and Photonics
• /
• 제5권6호
• /
• pp.686-698
• /
• 2021

A camouflaged encryption scheme based on Hadamard matrix and ghost imaging is proposed. In the process of the encryption, an orthogonal matrix is used as the projection pattern of ghost imaging to improve the definition of the reconstructed images. The ciphertext of the secret image is constrained to the camouflaged image. The key of the camouflaged image is obtained by the method of sparse decomposition by principal component orthogonal basis and the constrained ciphertext. The information of the secret image is hidden into the information of the camouflaged image which can improve the security of the system. In the decryption process, the authorized user needs to extract the key of the secret image according to the obtained random sequences. The real encrypted information can be obtained. Otherwise, the obtained image is the camouflaged image. In order to verify the feasibility, security and robustness of the encryption system, binary images and gray-scale images are selected for simulation and experiment. The results show that the proposed encryption system simplifies the calculation process, and also improves the definition of the reconstructed images and the security of the encryption system.

• 한국멀티미디어학회논문지
• /
• 제22권4호
• /
• pp.443-454
• /
• 2019

In recent years, there has been active research on a recommender system that considers three or more inputs in addition to users and goods, making it a multi-dimensional array, also known as a tensor. The main issue with using tensor is that there are a lot of missing values, making it sparse. In order to solve this, the tensor can be shrunk using the tensor decomposition algorithm into a lower dimensional array called a factor matrix. Then, the tensor is reconstructed by calculating factor matrices to fill original empty cells with predicted values. This is called tensor reconstruction. In this paper, we propose a user-based Top-K recommender system by normalized PARAFAC tensor reconstruction. This method involves factorization of a tensor into factor matrices and reconstructs the tensor again. Before decomposition, the original tensor is normalized based on each dimension to reduce overfitting. Using the real world dataset, this paper shows the processing of a large amount of data and implements a recommender system based on Apache Spark. In addition, this study has confirmed that the recommender performance is improved through normalization of the tensor.

• 대한전자공학회논문지SP
• /
• 제37권6호
• /
• pp.48-59
• /
• 2000

안테나 센서 어레이를 이용하여 수신되는 전파의 도래각을 추정하는 방식으로서 MUSIC(multiple signal classification)과 같은 고유분해(eigendecomposition)를 기반으로 한 방식은 백색잡음 환경하에서는 고분해능의 우수한 성능을 보이지만 유색잡음이 존재하는 환경에서는 성능이 크게 저하된다. 본 논문에서는 주기성을 가진 신호에 잡음이 더해진 선호를 웨이브렛 영역으로 변환하여 신호와 잡음을 분리하는 방법을 사용하여 유색잡음이 있는 환경에서 도래각 추정 문제를 접근하였다. 배경잡음만 있는 경우 센서 어레이 출력을 이산 웨이브렛 분해를 하여 얻은 멀티스케일 성분들의 공분산 행렬은 밴드화된 행렬로 근사화 할 수 있는데 비하여 협대역 신호는 멀티스케일 성분간의 상관성은 급속히 감소하는 현상을 보이지 않고 공분산 행렬에서는 신호성분이 전체 행렬에 분포한다. 어레이 출력의 공분산 행렬을 웨이브렛 영역으로 변환하여 유색잡음에 해당하는 특정 밴드를 삭제하고 MUSIC과 같은 기존의 공간 스펙트럼 추정방식을 적용하여 도래각을 추정 한 다음 그 결과로 부터 신호성분을 합성하여 삭제한 밴드를 채우는 과정을 반복하여 정확한 도래각을 얻는 방안을 제안하였다. 제안된 알고리즘의 성능을 여러 가지 형태의 상관함수 특성을 가진 유색잡음 환경에서 모의실험을 통하여 기존 방식과 비교 분석하였다.

• 전기학회논문지
• /
• 제66권7호
• /
• pp.1078-1082
• /
• 2017

This paper presents an L1 optimization based filtering technique which effectively eliminates the optical misalignment effects encountered in the squint guidance mode with strapdown seekers. We formulated a series of L1 optimization problems in order to separate the bias and the gradient components from the measured data, and solved them via the alternating direction method of multipliers (ADMM) and sparse matrix decomposition techniques. The proposed technique was able to rapidly detect arbitrary discontinuities and gradient changes from the measured signals, and was shown to effectively cancel the undesirable effects coming from the seeker misalignment angles. The technique was implemented on embedded flight computers and the real-time operational performance was verified via the hardware-in-the-loop simulation (HILS) tests in parallel with the automatic target recognition algorithms and the intra-red synthetic target images.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.352-368
• /
• 2021

Sign-in point of interest (POI) are extremely sparse in location-based social networks, hindering recommendation systems from capturing users' deep-level preferences. To solve this problem, we propose a content-aware POI recommendation algorithm based on a convolutional neural network. First, using convolutional neural networks to process comment text information, we model location POI and user latent factors. Subsequently, the objective function is constructed by fusing users' geographical information and obtaining the emotional category information. In addition, the objective function comprises matrix decomposition and maximisation of the probability objective function. Finally, we solve the objective function efficiently. The prediction rate and F1 value on the Instagram-NewYork dataset are 78.32% and 76.37%, respectively, and those on the Instagram-Chicago dataset are 85.16% and 83.29%, respectively. Comparative experiments show that the proposed method can obtain a higher precision rate than several other newer recommended methods.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1989년도 하계종합학술대회 논문집
• /
• pp.202-206
• /
• 1989

A new algorithm is suggested to solve the optimal reactive power control(optimal VAR control) problem. An efficient computer program based on the latest achievements in the sparse matrix/vector techniques has been developed for this purpose. The model minimizes the real power losses in the system. The constraints include the reactive power limits of the generators, limits on the bus voltages and the operating limits of control variables- the transformer tap positions, generator terminal voltages and switchable reactive power sources. The method developed herein employs linearized sensitivity relationships of power systems to establish both the objective function for minimizing the system losses and the system performance sensitivities relating dependent and control variables. The algorithm consists of two modules, i.e. the Q-V module for reactive power-voltage control, Load flow module for computational error adjustments. In particular, the acceleration factor technique is introduced to enhance the convergence property in Q-module, The combined use of the afore-mentioned two modules ensures more effective and efficient solutions for optimal reactive power dispatch problems. Results of the application of the method to the sample system and other worst-case system demonstrated that the algorithm suggested herein is compared favourably with conventional ones in terms of computation accuracy and convergence characteristics.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권12호
• /
• pp.3798-3814
• /
• 2022

Social recommendation algorithm can alleviate data sparsity and cold start problems in recommendation system by integrated social information. Among them, matrix-based decomposition algorithms are the most widely used and studied. Such algorithms use dot product operations to calculate the similarity between users and items, which ignores user's potential preferences, reduces algorithms' recommendation accuracy. This deficiency can be avoided by a metric learning-based social recommendation algorithm, which learns the distance between user embedding vectors and item embedding vectors instead of vector dot-product operations. However, previous works provide no theoretical explanation for its plausibility. Moreover, most works focus on the indirect impact of social friends on user's preferences, ignoring the direct impact on user's rating preferences, which is the influence of user rating preferences. To solve these problems, this study proposes a user bias drift social recommendation algorithm based on metric learning (BDML). The main work of this paper is as follows: (1) the process of introducing metric learning in the social recommendation scenario is introduced in the form of equations, and explained the reason why metric learning can replace the click operation; (2) a new user bias is constructed to simultaneously model the impact of social relationships on user's ratings preferences and user's preferences; Experimental results on two datasets show that the BDML algorithm proposed in this study has better recommendation accuracy compared with other comparison algorithms, and will be able to guarantee the recommendation effect in a more sparse dataset.