• 인터넷정보학회논문지
• /
• 제8권6호
• /
• pp.95-102
• /
• 2007

협력적 필터링을 이용한 추천시스템은 희소성의 문제로 인해 예측의 정확도에 대한 신뢰성에 문제가 있다. 이는 선호도 평가치의 희소성이 크면 이웃선정과정에 문제가 있을 뿐만 아니라 예측의 정확도를 떨어뜨린다. 본 논문에서는 사용자의 응답 희소성에 따른 MAE의 변화를 조사하였으며 희소성에 따라 집단을 분류하고 분류된 집단에 따른 MAE는 유의적인 차이가 있는 지를 분석하였다. 그리고 희소성 문제로 인한 집단 간의 예측 정확도를 높이기 위한 방법으로 희소성이 있는 아이템을 선별하여 이들 중에서 선호도 응답이 많은 사용자 고객의 선호도 평균값을 선호도 평가 치로 대치시켜 희소성을 완화하여 추천시스템의 예측 정확도가 높아졌음을 연구하였다.

• Asia pacific journal of information systems
• /
• 제14권2호
• /
• pp.131-149
• /
• 2004

Collaborative filtering is one of popular techniques for personalized recommendation in e-commerce sites. An advantage of collaborative filtering is that the technique can work with sparse evaluation data to predict preference scores of new alternative contents or advertisements. There is, however, no in-depth study about the sparsity effect of customer's evaluation data to the performance of recommendation. In this study, we investigate the sparsity effect and hybrid usages of customers' evaluation data and purchase data using an experiment result. The result of the analysis shows that the performance of recommendation decreases monotonically as the sparsity increases, and also the hybrid usage of two different types of data; customers' evaluation data and purchase data helps to increase the performance of recommendation in sparsity situation.

• Management Science and Financial Engineering
• /
• 제11권3호
• /
• pp.19-45
• /
• 2005

Collaborative filtering is one of the most common methods that e-commerce sites and Internet information services use to personalize recommendations. Collaborative filtering has the advantage of being able to use even sparse evaluation data to predict preference scores for new products. To date, however, no in-depth investigation has been conducted on how the data sparsity effect in customers' evaluation data affects collaborative filtering-based recommendation performance. In this study, we analyzed the sparsity effect and used a hybrid method based on customers' evaluations and purchases collected from an online bookstore. Results indicated that recommendation performance decreased monotonically as sparsity increased, and that performance was more sensitive to sparsity in evaluation data rather than in purchase data. Results also indicated that the hybrid use of two different types of data (customers' evaluations and purchases) helped to improve the recommendation performance when evaluation data were highly sparse.

• Current Optics and Photonics
• /
• 제5권4호
• /
• pp.431-443
• /
• 2021

Sparse unmixing has been proven to be an effective method for hyperspectral unmixing. Hyperspectral images contain rich spectral and spatial information. The means to make full use of spectral information, spatial information, and enhanced sparsity constraints are the main research directions to improve the accuracy of sparse unmixing. However, many algorithms only focus on one or two of these factors, because it is difficult to construct an unmixing model that considers all three factors. To address this issue, a novel algorithm called multiview-based spectral weighted and low-rank row-sparsity unmixing is proposed. A multiview data set is generated through spectral partitioning, and then spectral weighting is imposed on it to exploit the abundant spectral information. The row-sparsity approach, which controls the sparsity by the l_2,0 norm, outperforms the single-sparsity approach in many scenarios. Many algorithms use convex relaxation methods to solve the l_2,0 norm to avoid the NP-hard problem, but this will reduce sparsity and unmixing accuracy. In this paper, a row-hard-threshold function is introduced to solve the l_2,0 norm directly, which guarantees the sparsity of the results. The high spatial correlation of hyperspectral images is associated with low column rank; therefore, the low-rank constraint is adopted to utilize spatial information. Experiments with simulated and real data prove that the proposed algorithm can obtain better unmixing results.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권1호
• /
• pp.58-78
• /
• 2016

Compressed sensing (CS) possesses the potential benefits for spectrum sensing of wideband signal in cognitive radio. The sparsity of signal in frequency domain denotes the number of occupied channels for spectrum sensing. This paper presents a scheme of adaptively adjusting the number of compressed measurements to reduce the unnecessary computational complexity when priori information about the sparsity of signal cannot be acquired. Firstly, a method of sparsity estimation is introduced because the sparsity of signal is not available in some cognitive radio environments, and the relationship between the amount of used data and estimation accuracy is discussed. Then the SNR of the compressed signal is derived in the closed form. Based on the SNR of the compressed signal and estimated sparsity, an adaptive algorithm of adjusting the number of compressed measurements is proposed. Finally, some simulations are performed, and the results illustrate that the simulations agree with theoretical analysis, which prove the effectiveness of the proposed adaptive adjusting of compressed measurements.

• Journal of Communications and Networks
• /
• 제15권3호
• /
• pp.283-291
• /
• 2013

In this paper, we propose new multiuser detectors (MUDs) based on compressed sensing approaches for the large-scale multiple antenna systems equipped with dozens of low-power antennas. We consider the scenarios where the number of receiver antennas is smaller than the total number of users, but the number of active users is relatively small. This prior information motivates sparsity-embracing MUDs such as sparsity-embracing linear/nonlinear MUDs where the detection of active users and their symbol detection are employed. In addition, sparsity-embracing MUDs with maximum a posteriori probability criterion (MAP-MUDs) are presented. They jointly detect active users and their symbols by exploiting the probability of user activity, and it can be solved efficiently by introducing convex relaxing senses. Furthermore, it is shown that sparsity-embracing MUDs exploiting common users' activity across multiple symbols, i.e., frame-by-frame, can be considered to improve performance. Also, in multiple multiple-input and multiple-output networks with aggressive frequency reuse, we propose the interference cancellation strategy for the proposed sparsity-embracing MUDs. That first cancels out the interference induced by adjacent networks and then recovers the desired users' information by exploiting the low user activity. In simulation studies for binary phase shift keying modulation, numerical evidences establish the effectiveness of our proposed MUDs exploiting low user activity, as compared with the conventional MUD.

• 한국통신학회논문지
• /
• 제40권7호
• /
• pp.1234-1238
• /
• 2015

본 논문에서는 Sub-Nyquist 샘플링 기반의 협력 스펙트럼 센싱 기법을 제안한다. 최근 압축 센싱 (Compressive Sensing) 기술이 많은 주목을 받으면서 원본 신호의 성긴 정도 (Sparsity)를 추정하는 기법도 활발히 연구되고 있다. 따라서 본 논문에서는 주파수 대역의 Sparsity를 안다고 가정할 때 다양한 샘플링율과 협력 센싱 기법에 따른 Sub-Nyquist 샘플링 기법의 검출 성능을 수학적으로 분석한다. 또한 모의실험 결과를 통해 제안된 기법의 성능을 입증한다.

• 대한전자공학회논문지SP
• /
• 제46권5호
• /
• pp.25-31
• /
• 2009

Compressed sensing은 기존의 Nyquist sampling 이론에 기반을 두었던 dynamic MRI에서의 시 공간 해상도의 제한을 획기적으로 향상시킴으로써, 최근 몇 년 사이, MR reconstruction 분야에서 가장 큰 이슈가 되고 있는 연구주제이다. Dynamic MRI 는 대부분 시간방향의 redundancy 가 매우 크므로, 쉽게 sparse 변환이 가능하다. 따라서 sparsity를 기본 조건으로 하는 compressed sensing은 거의 모든 dynamic MRI 에 대해 효과적으로 적용될 수 있다. 본 review 페이퍼에서는 최근 compressed sensing 에 기반을 두거나 영상의 sparsity를 이용하여 개발된 dynamic MR imaging algorithm 들을 간략히 소개하고, 비교 분석함으로써, compressed sensing과 같은 새로운 접근 방식의 dynamic MRI가 실제 임상에서 가져다 줄 발전 가능성을 제시한다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권8호
• /
• pp.3086-3101
• /
• 2021

To supply precise marketing and differentiated service for the electric power service department, it is very important to predict the customers with high sensitivity of electric power failure. To solve this problem, we propose a novel grouped 𝑙_1/2 sparsity constrained logistic regression method for sensitivity assessment of electric power failure. Different from the 𝑙₁ norm and k-support norm, the proposed grouped 𝑙_1/2 sparsity constrained logistic regression method simultaneously imposes the inter-class information and tighter approximation to the nonconvex 𝑙₀ sparsity to exploit multiple correlated attributions for prediction. Firstly, the attributes or factors for predicting the customer sensitivity of power failure are selected from customer sheets, such as customer information, electric consuming information, electrical bill, 95598 work sheet, power failure events, etc. Secondly, all these samples with attributes are clustered into several categories, and samples in the same category are assumed to be sharing similar properties. Then, 𝑙_1/2 norm constrained logistic regression model is built to predict the customer's sensitivity of power failure. Alternating direction of multipliers (ADMM) algorithm is finally employed to solve the problem by splitting it into several sub-problems effectively. Experimental results on power electrical dataset with about one million customer data from a province validate that the proposed method has a good prediction accuracy.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 1996년도 추계학술대회발표논문집; 고려대학교, 서울; 26 Oct. 1996
• /
• pp.289-292
• /
• 1996

In Cholesky factorization of the interior point method, dense columns of A matrix make dense Cholesky factor L regardless of sparsity of A matrix. We introduce a method to transform a primal problem to a dual problem in order to preserve the sparsity.