• Management Science and Financial Engineering
• /
• v.16 no.1
• /
• pp.81-117
• /
• 2010

Semi-supervised learning incorporates unlabeled examples, whose labels are unknown, as well as labeled examples into learning process. Although transductive support vector machine (TSVM), one of semi-supervised learning models, was proposed about a decade ago, its application to large-scaled data has still been limited due to its high computational complexity. Our previous research addressed this limitation by introducing a branch-and-bound algorithm for finding an optimal solution to TSVM. In this paper, we propose three new techniques to enhance the performance of the branch-and-bound algorithm. The first one tightens min-cut bound, one of two bounding strategies. Another technique exploits a graph-based approximation to a support vector machine problem to avoid the most time-consuming step. The last one tries to fix the labels of unlabeled examples whose labels can be obviously predicted based on labeled examples. Experimental results are presented which demonstrate that the proposed techniques can reduce drastically the number of subproblems and eventually computational time.

• Journal of Power System Engineering
• /
• v.9 no.4
• /
• pp.202-208
• /
• 2005

The genetic algorithm (GA) which is one of the popular optimum algorithm has been used to solve a variety of optimum problems. Because it need not require the gradient of objective function and is easier to find global solution than gradient-based optimum algorithm using the gradient of objective function. However optimum method using the GA and the finite element method (FEM) takes many computational time to solve the optimum structural design problem which has a great number of design variables, constraints, and system with many degrees of freedom. In order to overcome the drawback of the optimum structural design using the GA and the FEM, the author developed a computer program which can optimize frame structures by using the GA and the generalized transfer stiffness coefficient method. In order to confirm the effectiveness of the developed program, it is applied to optimum design of plane frame structures. The computational results by the developed program were compared with those of iterative design.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.11
• /
• pp.4573-4584
• /
• 2015

The traditional feature extraction algorithms for recognition of communication signals can hardly realize the balance between computational complexity and signals' interclass gathered degrees. They can hardly achieve high recognition rate at low SNR conditions. To solve this problem, a novel feature extraction algorithm based on Holder coefficient was proposed, which has the advantages of low computational complexity and good interclass gathered degree even at low SNR conditions. In this research, the selection methods of parameters and distribution properties of the extracted features regarding Holder coefficient theory were firstly explored, and then interval gray relation algorithm with improved adaptive weight was adopted to verify the effectiveness of the extracted features. Compared with traditional algorithms, the proposed algorithm can more accurately recognize signals at low SNR conditions. Simulation results show that Holder coefficient based features are stable and have good interclass gathered degree, and interval gray relation classifier with adaptive weight can achieve the recognition rate up to 87% even at the SNR of -5dB.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1998.04a
• /
• pp.216-225
• /
• 1998

Although artificial neural network based on backpropagation algorithm is an excellent system simulator, it has still unsolved problems of its structure-decision and learning method. That is, we cannot find a general approach to decide the structure of the neural network and cannot train it satisfactorily because of the local optimum point which it frequently falls into. In addition, although there are many successful applications using backpropagation learning algorithm, there are few efforts to improve the learning algorithm itself. In this study, we suggest a general way to construct the hidden layer of the neural network using binary genetic algorithm and also propose the various learning methods by which the global minimum value of the teaming error can be obtained. A XOR problem and line heating problems are investigated as examples.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2000.10a
• /
• pp.315-322
• /
• 2000

In this study, an algorithm analyzing dynamic mutiple-crack propagation problem by Meshfree Method is proposed. A short description of Meshfree Method especially, Element-free Galerkin (EFG) method is presented and the elastodynamic fracture theory is summarized. A numerical implementation algorithm for dynamic analysis by Meshfree Method is discussed and an algorithm for mutlple-crack dynamic propagation is also presented. A couple of numerical examples of dynamic crack propagation problem illustrate the performance of the proposed technique. The accuracy of the algorithm is studied in the first example by being compared with experimental results, and the applicability and efficiency of the developed algorithm is studied in the second example.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2000.10a
• /
• pp.307-314
• /
• 2000

A parallel condensation algorithm for efficient dynamic analysis of three-dimensional large-scale structures is presented. The algorithm is developed for a user-friendly and cost effective high-performance computing system on a collection of Pentium processors connected via a 100 Mb/s Ethernet LAN. To harness the parallelism in the computing system effectively, a large-scale structure is partitioned into a number of substructures equal to the number of computers in the computing system Then, for reduction in the size of an eigenvalue problem the computations required for static condensation of each substructure is processed concurrently on each slave computer. The performance of th proposed parallel algorithm is demonstrated by applying to dynamic analysis of a three dimensional structure. The results show that how the parallel algorithm facilitates the efficient use of a small number of low-cost personal computers for dynamic analysis of large-scale structures.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2003.10a
• /
• pp.115-122
• /
• 2003

A Metropolis genetic algorithm(MGA) is developed and applied for the structural design optimization. In MGA favorable features of Metropolis algorithm in simulated annealing(SA) are incorporated in simple genetic algorithm(SGA), so that the MGA alleviates the disadvantage of finding imprecise solution in SGA and time-consuming computation in SA. Performances of MGA are compared with those of conventional algorithms such as Holland's SGA, Krishnakumar's micro genetic algorithm(μGA), and Kirkpatrick's SA. Typical numerical examples are used to evaluate the favorable features and applicability of MGA From the theoretical evaluation and numerical experience, it is concluded that the proposed MGA is a reliable and efficient tool for structural design optimization.

• Journal of computational fluids engineering
• /
• v.7 no.1
• /
• pp.1-9
• /
• 2002

In this paper, a modification of PISO algorithm based on standard k-ε turbulence model was proposed. The numerical technique used in this research is finite volume method, hybrid scheme for discretizing convection term, Euler implicit scheme for discretizing time term, and non-staggered grid. The basic idea of the modification of PISO algorithm is to perform an additional corrector stage for turbulence kinetic energy and dissipation rate to correct the inconsistence of flow and turbulence. In order to validate this algorithm, simulation of flow around a square cylinder (Re=3000) was performed in two-dimensional case. The results obtained from the proposed scheme show better agreement with those from the experiment than using original PISO algorithm in coherent velocity field.

• Korean Journal of Computational Design and Engineering
• /
• v.7 no.2
• /
• pp.75-80
• /
• 2002

This paper details a robust and efficient algorithm for point classification with respect to a polygon in 2D real number domain. The concept of tolerance makes this algorithm robust and consistent. It enables to define‘on-boundary’ , which can be interpreted as either‘in-’or‘out-’side region, and to manage rounding errors in floating point computation. Also the tolerance is used as a measure of reliability of point classifications. The proposed algorithm is based on a ray-intersection technique known as the most efficient, in which intersections between a ray originating from a given test point and the boundary of a region are counted. An odd number of intersections indicates that the point is inside region. For practical examples the algorithm is most efficient because most edges of the polygon region are processed by simple bit operations.

• Korean Journal of Computational Design and Engineering
• /
• v.22 no.2
• /
• pp.110-117
• /
• 2017

Delaunay refinement algorithm is a classical method to generate quality triangular meshes when point cloud and/or constrained edges are given in two- or three-dimensional space. It computes the Delaunay triangulation for given points and edges to obtain an initial solution, and update the triangulation by inserting steiner points one by one to get an improved quality triangulation. This process repeats until it satisfies given quality criteria. The efficiency of the algorithm depends on the criteria and point insertion method. In this paper, we propose a method to accelerate the Delaunay refinement algorithm by applying geometric hashing technique called bucketing when inserting a new steiner point so that it can localize necessary computation. We have tested the proposed method with a few types of data sets, and the experimental result shows strong linear time behavior.