• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.532-534
• /
• 2021

최근 데이터의 양이 급격히 증가함에 따라 빅데이터 환경에서 데이터 질의 처리 수행 시 연산 시간이 많이 소요되는 문제점이 발생한다. 이러한 처리 시간을 줄이기 위한 방법으로 근사질의 처리에 대한 연구의 필요성이 대두되고 있다. 근사 질의 처리 방법은 정확도가 다소 떨어지더라도 빠른 결과를 요구하는 응용 분야에서 매우 유용하게 쓰일 수 있다. 본 논문에서는 사용자가 원하는 결과 정확도와 적시성 등을 지원하기 위한 근사 질의 처리 언어 확장, 실행 계획생성 및 질의 최적화 기술을 제안하고, 설계 방향 및 특징 등에 대해서 설명한다.

• Proceedings of the Korean Nuclear Society Conference
• /
• 2004.10a
• /
• pp.953-954
• /
• 2004

• Proceedings of the Korea Air Pollution Research Association Conference
• /
• 2010.04a
• /
• pp.594-595
• /
• 2010

• Korean Journal of Optics and Photonics
• /
• v.17 no.4
• /
• pp.359-365
• /
• 2006

Wave-front error(WFE) is the main parameter that determines the optical performance of the opto-mechanical system. In the development of opto-mechanics, WFE due to the main loading conditions are set to the important specifications. The deformation of the optical surface can be exactly calculated thanks to the evolution of numerical methods such as the finite element method(FEM). To calculate WFE from the deformation results of FEM, another approximation of the optical surface deformation is required. It needs to construct additional grid or element mesh. To construct additional mesh is troublesomeand leads to transformation error. In this work, the moving least-squares approximation is used to reconstruct wave front error It has the advantage of accurate approximation with only nodal data. There is no need to construct additional mesh for approximation. The proposed method is applied to the examples of GOCI scan mirror in various loading conditions. The validity is demonstrated through examples.

• Journal of KIISE:Software and Applications
• /
• v.28 no.9
• /
• pp.681-687
• /
• 2001

In order to raise a class discrimination power by combining multiple classifiers under the Bayesian decision theory, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables obtained from training data samples should be minimized. Wang and Wong proposed a tree dependence first-order approximation scheme of a high order probability distribution composed of the class and multiple feature pattern variables for minimizing the upper bound of the Bayes error rate. This paper presents an extended high order product approximation scheme dealing with higher order dependency more than the first-order tree dependence, based on the minimization of the upper bound of the Bayes error rate. Multiple recognizers for unconstrained handwritten numerals from CENPARMI were combined by the proposed approximation scheme using the Bayesian formalism, and the high recognition rates were obtained by them.

• The Journal of the Korea Contents Association
• /
• v.20 no.8
• /
• pp.42-53
• /
• 2020

With the development of social network services, graph structures have been utilized to represent relationships among objects in various applications. Recently, a demand of subgraph matching in real-time graph streams has been increased. Therefore, an efficient approximate Top-k subgraph matching scheme for low latency in real-time graph streams is required. In this paper, we propose an approximate Top-k subgraph matching scheme considering data reuse in graph stream environments. The proposed scheme utilizes the distributed stream processing platform, called Storm to handle a large amount of stream data. We also utilize an existing data reuse scheme to decrease stream processing costs. We propose a distance based summary indexing technique to generate Top-k subgraph matching results. The proposed summary indexing technique costs very low since it only stores distances among vertices that are selected in advance. Finally, we provide k subgraph matching results to users by performing an approximate Top-k matching on the summary indexing. In order to show the superiority of the proposed scheme, we conduct various performance evaluations in diverse real world datasets.

• The KIPS Transactions:PartA
• /
• v.16A no.6
• /
• pp.509-518
• /
• 2009

In this paper, a mechanism that produces an approximation edges minimum spanning tree swiftly using virtual nodes called portals dividing given edges into same distance sub-edges. The approximation edges minimum spanning tree can be used in many useful areas as connecting communication lines, road networks and railroad systems. For 3000 random input edges, when portal distance is 0.3, tree building time decreased 29.74% while the length of the produced tree increased 1.8% comparing with optimal edge minimum spanning tree in our experiment. When portal distance is 0.75, tree building time decreased 39.96% while the tree length increased 2.96%. The result shows this mechanism might be well applied to the applications that may allow a little length overhead, but should produce an edge connecting tree in short time. And the proposed mechanism can produce an approximation edge minimum spanning tree focusing on tree length or on building time to meet user requests by adjusting portal distance or portal discard ratio as parameter.

• Journal of Korean Society of Steel Construction
• /
• v.12 no.1 s.44
• /
• pp.45-54
• /
• 2000

Design-oriented approximate structural reanalysis models are compared and assessed, particularly with focus on the case of large changes of design variables. The effectiveness and reliability are demonstrated by means of numerical examples. The results of the study suggest the following conclusions relative to the potential of the procedures. (A) local approximation is only appropriate for the case of small changes in design : (B) global approximation is exact for the case of large changes in a small number of design variables, but inefficient : (C) local-global approximation is most effective and reliable for the case of large changes with a large number of design variables. These methods can improve the total efficiency when they are appropriately used to the design information for the redesign process of large scale structures.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.10
• /
• pp.2303-2308
• /
• 2012

In this paper, we deal with the problem of scheduling parallel tasks with deadlines. Parallel tasks can be simultaneously executed on various machines and specially, we consider the malleable tasks, that is, the tasks whose execution time is given by a function of the number of machines on which they are executed. The goal of the problem is to maximize the throughput of tasks completed within their deadlines. This problem is well-known as NP-hard problem. Thus we will find an approximation algorithm, and its performance is compared with that of the optimal algorithm and analyzed by finding the approximation ratio. In particular, the algorithm has more resources, that is, more machines, than the optimal algorithm. This is called the resource augmentation analysis. We propose an algorithm to guarantee the approximation ratio of 3.67 using 1.5 times machines.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.5
• /
• pp.98-106
• /
• 1998

This paper proposes an efficient learning algorithm for improving the training performance of the neural network. The proposed method improves the training performance by applying the backpropagation algorithm of a global optimization method which is a hybrid of a stochastic approximation and a conjugate gradient method. The approximate initial point for f a ~gtl obal optimization is estimated first by applying the stochastic approximation, and then the conjugate gradient method, which is the fast gradient descent method, is applied for a high speed optimization. The proposed method has been applied to the parity checking and the pattern classification, and the simulation results show that the performance of the proposed method is superior to those of the conventional backpropagation and the backpropagation algorithm which is a hyhrid of the stochastic approximation and steepest descent method.