• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.5
• /
• pp.2607-2627
• /
• 2017

Nonnegative matrix factorization (NMF) has received considerable attention due to its effectiveness of reducing high dimensional data and importance of producing a parts-based image representation. Most of existing NMF variants attempt to address the assertion that the observed data distribute on a nonlinear low-dimensional manifold. However, recent research results showed that not only the observed data but also the features lie on the low-dimensional manifolds. In addition, a few hard priori label information is available and thus helps to uncover the intrinsic geometrical and discriminative structures of the data space. Motivated by the two aspects above mentioned, we propose a novel algorithm to enhance the effectiveness of image representation, called Dual graph-regularized Constrained Nonnegative Matrix Factorization (DCNMF). The underlying philosophy of the proposed method is that it not only considers the geometric structures of the data manifold and the feature manifold simultaneously, but also mines valuable information from a few known labeled examples. These schemes will improve the performance of image representation and thus enhance the effectiveness of image classification. Extensive experiments on common benchmarks demonstrated that DCNMF has its superiority in image classification compared with state-of-the-art methods.

• Journal of the Korean Society for Precision Engineering
• /
• v.25 no.7
• /
• pp.103-110
• /
• 2008

This paper presents a new approach of recognizing a 3D object using a single camera, based on the extended convex hull of its silhouette. It aims at minimizing the DB size and simplifying the processes for matching and feature extraction. For this purpose, two concepts are introduced: extended convex hull and measurable region. Extended convex hull consists of convex curved edges as well as convex polygons. Measurable region is the cluster of the viewing vectors of a camera represented as the points on the orientation sphere from which a specific set of surfaces can be measured. A measurable region is represented by the extended convex hull of the silhouette which can be obtained by viewing the object from the center of the measurable region. Each silhouette is represented by a relation graph where a node describes an edge using its type, length, reality, and components. Experimental results are included to show that the proposed algorithm works efficiently even when the objects are overlapped and partially occluded. The time complexity for searching the object model in the database is O(N) where N is the number of silhouette models.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.39 no.2
• /
• pp.131-140
• /
• 2014

Customer networks go through constant changes. They may expand or shrink once they are formed. In dynamic environments, it is a critical corporate challenge to identify and manage influential customer groups in a cost effective way. In this context, we apply inverse optimization theory to suggest an efficient method to manage customer networks. In this paper, we assume that there exists a subset of nodes that might have a large effect on the network and that the network can be modified via some strategic actions. Rather than making efforts to find influential nodes whenever the network changes, we focus on a subset of selective nodes and perturb as little as possible the interaction between nodes in order to make the selected nodes influential in the given network. We define the following problem based on the inverse optimization. Given a graph and a prescribed node subset, the objective is to modify the structure of the given graph so that the fixed subset of nodes becomes a minimum dominating set in the modified graph and the cost for modification is minimum under a fixed norm. We call this problem the inverse dominating set problem and investigate its computational complexity.

• The KIPS Transactions:PartB
• /
• v.8B no.6
• /
• pp.651-656
• /
• 2001

In this paper, we propose Improved Self-Organized Graph(Improved SOG：$SOG^*$)thinning method, which maintains the excellent thinning results of Self-organized graph(SOG) built from Self-Organizing features map and improves the performance of modified SOG using a new incremental learning method of Kohonen features map. In the experiments, this method shows the thinning results equal to those of SOG and the time complexity O((logM)3) superior to it. Therefore, the proposed method is useful for the feature extraction from digits and characters in the preprocessing step.

• Journal of the Korean Ceramic Society
• /
• v.53 no.5
• /
• pp.521-528
• /
• 2016

Kinetic Monte Carlo (KMC) and graph searches show that proton conduction limiting barriers and trajectories in $BaZr_{0.875}Y_{0.125}O_3$ are affected by the presence of other protons. At 1000 K, KMC limiting conduction barriers increase from 0.39 eV to 0.45 eV as the proton number is increased. The proton-proton radial distribution begins to rise at $2{\AA}$ and peaks at $4{\AA}$, which is half the distance expected, based on the proton concentration. Density functional theory (DFT) calculations find proton/proton distances of 2.60 and $2.16{\AA}$ in the lowest energy two-proton configurations. A simple average of the limiting barriers for 7-10 step periodic long range paths found via graph theory at 1100 K shows an increase in activation barrier from 0.32 eV to 0.37 eV when a proton is added. Both KMC and graph theory show that protons can affect each other's pathways and raise the overall conduction barriers.

• Nuclear Engineering and Technology
• /
• v.48 no.2
• /
• pp.386-403
• /
• 2016

The safety of nuclear power plants is analyzed by a probabilistic risk assessment, and the fault tree analysis is the most widely used method for a risk assessment with the event tree analysis. One of the well-known disadvantages of the fault tree is that drawing a fault tree for a complex system is a very cumbersome task. Thus, several graphical modeling methods have been proposed for the convenient and intuitive modeling of complex systems. In this paper, the reliability graph with general gates (RGGG) method, one of the intuitive graphical modeling methods based on Bayesian networks, is improved for the reliability analyses of dynamic systems that have various operation modes with time. A reliability matrix is proposed and it is explained how to utilize the reliability matrix in the RGGG for various cases of operation mode changes. The proposed RGGG with a reliability matrix provides a convenient and intuitive modeling of various operation modes of complex systems, and can also be utilized with dynamic nodes that analyze the failure sequences of subcomponents. The combinatorial use of a reliability matrix with dynamic nodes is illustrated through an application to a shutdown cooling system in a nuclear power plant.

• Tunnel and Underground Space
• /
• v.18 no.4
• /
• pp.243-251
• /
• 2008

Recently, the mining methods are changing from surface mining to underground mining because of the increment of the environmental issues and legal regulations. Therefore, the stability of underground openings is a major concern for the safety and productivity of mining operations. In this paper, a survey of structural geology and discontinuities were carried out at a limestone mine. The relevant mechanical properties of rocks were determined by the laboratory tests and rock mass classifications (RMR and Q-system) for the mine design and input data for the stability analysis. The dimensions of unsupported span for underground openings and pillar were decided based on the RMR values of rock mass classifications. The stability analysis for the suggested mine design was examined through the empirical methods (stability graph method and critical span curve) and 3-D numerical analysis (Visual-FEA).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.5
• /
• pp.2211-2232
• /
• 2018

The unstructured and semi-structured big data in social network poses new challenges in query retrieval. This requirement needs to be met by introducing quality retrieval time measures like indexing. Due to the huge volume of data storage, there originate the need for efficient index algorithms to promote query processing. However, conventional algorithms fail to index the huge amount of frequently obtained information in real time and fall short of providing scalable indexing service. In this paper, a new LIndex algorithm, which is a heuristic on Lucene is built on Neo4jHA architecture that holds the social network Big data. LIndex is a flexible and simplified adaptive indexing scheme that ascendancy decomposed shortest paths around term neighbors as basic indexing unit. This newfangled index proves to be effectual in query space pruning of graph database Neo4j, scalable in index construction and deployment. A graph query is processed and optimized beyond the traditional Lucene in a time-based manner to a more efficient path method in LIndex. This advanced algorithm significantly reduces query fetch without compromising the quality of results in time. The experiments are conducted to confirm the efficiency of the proposed query retrieval in Neo4j graph NoSQL database.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.38 no.6
• /
• pp.831-838
• /
• 2018

This paper presents a method to estimate the flow discharge in a backwater affected river junction. First, unsteady HEC-RAS model was simulated and calibrated using 2 recent real flood and then HPG (Hydraulic Performance Graph) was created by plotting the relationship between upstream and downstream stages and discharge in the reach and performing kriging interpolation. During a flood, the discharge through the reach can be estimated based on the stages at its ends and the developed HPG. These discharge data were in good agreement with the automatic discharge measurements such as ADVM. This study could provide an economical and practical method for estimating discharge in a junction with a high hysteresis of stage-discharge relationships.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.275-278
• /
• 2008

Recently, there are many research of datamining. On the transaction dataset, association rules is made by finding of interesting patterns. A part of mining, sub-structure mining is increased in interest of and applied to many high technology. But graph mining has more computing time then itemset mining. Therefore, that need efficient way for avoid duplication. GASTON is best algorithm of duplication free. This paper analyze GASTON and expect the future work.