• MALSORI
• /
• no.66
• /
• pp.105-116
• /
• 2008

In this research, we propose a speaker identification system using a kernel method which is expected to model the non-linearity of speech features well. We have been using principal component analysis (PCA) successfully, and extended to kernel PCA, which is used for many pattern recognition tasks such as face recognition. However, we cannot use kernel PCA for speaker identification directly because the storage required for the kernel matrix grows quadratically, and the computational cost grows linearly (computing eigenvector of $l{\times}l$ matrix) with the number of training vectors I. Therefore, we use greedy kernel PCA which can approximate kernel PCA with small representation error. In the experiments, we compare the accuracy of the greedy kernel PCA with the baseline Gaussian mixture models using MFCCs and PCA. As the results with limited enrollment data show, the greedy kernel PCA outperforms conventional methods.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.4
• /
• pp.1-8
• /
• 2024

The traveling salesman problem(TSP) is one of the well known combinatorial optimization problems. Local search has been used as a method to solve TSP. Greedy Random Insertion(GRI) is known as an effective neighbor generation method for local search. GRI selects some cities from the current solution randomly and inserts them one by one into the best position of the current partial solution considering only one city at a time. We first propose another greedy neighbor generation method which is named Full Greedy Insertion(FGI). FGI determines insertion location one by one like GRI, but considers all remaining cities at once. And then we propose a method to combine GRI with FGI, in which GRI or FGI is randomly selected and executed at each iteration in simulated annealing. According to the experimental results, FGI alone does not necessarily perform very well. However, we confirmed that the combined method outperforms the existing local search methods including GRI.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.13 no.1
• /
• pp.79-85
• /
• 2017

Vehicular ad-hoc networks (VANETs) cause link disconnection problems due to the rapid speed and the frequent moving direction change of vehicles. Link disconnection in vehicle-to-vehicle communication is an important issue that must be solved because it decreases the reliability of packet forwarding. From the characteristics of VANETs, greedy forwarding protocols using the position information based on the inter-vehicle distance have gained attention. However, greedy forwarding protocols do not perform well in the urban environment where the direction of the vehicle changes greatly. It is because greedy forwarding protocols select the neighbor vehicle that is closest to the destination vehicle as the next transmission vehicle. In this paper, we propose a greedy anycast forwarding (GAF) protocol to improve the reliability of the inter-vehicle communication. The proposed GAF protocol combines the greedy forwarding scheme and the anycast forwarding method. Simulation results show that the GAF protocol can provide a better packet delivery rate than existing greedy forwarding protocols.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.7
• /
• pp.62-71
• /
• 1998

In this paper, we describe the object-oriented motioncompensation technique using the hierarchical grid interpolation(HGI). By noting that the HGI does not exploit effetively the rate-distortion (R-D) trade-off inherent in the quadtree decomposition, we propose an objectoriented HGI technique employing the greedy method. In the proposed technique, input image is decomposed in a quadtree basis using the greedy method, yielding maximum split gain in the R-D sense. Then, the motion compensation is performed using the HGI technique. The performance of the proposed technique is examined in simulation, and it will be show that the proposed technique provides better performance than the conventional object-oriented motion compensation techniques.

• The KIPS Transactions:PartB
• /
• v.14B no.7
• /
• pp.541-548
• /
• 2007

Due to the mathematical convergence property, Simulated Annealing (SA) has been one of the most popular optimization algorithms. However, because of its problem of slow convergence in the practical use, many variations of SA like Fast SA (FSA) have been developed for faster convergence. In this paper, we propose and prove that Greedy SA (GSA) also finds the global optimum in probability in the continuous space optimization problems. Because the greedy selection does not allow the cost to become worse, GSA is expected to have faster convergence than the conventional FSA that uses Metropolis selection. In the computer simulation, the proposed method is shown to have as good performance as FSA with Metropolis selection in the viewpoints of the convergence speed and the quality of the found solution. Furthermore, the greedy selection does not concern the cost value itself but uses only dominance of the costs of solutions, which makes GSA invariant to the problem scaling.

• Journal of the military operations research society of Korea
• /
• v.13 no.1
• /
• pp.101-115
• /
• 1987

This paper presents an optimization based heuristic algorithm for a tour bus scheduling problem where buses consist of various kinds of sightseeing and commutation services. First, this algorithm transforms the prolem into a vehicle routing problem on whose nodes denote trips and arcs denote connections between trips. Second, a greedy heuritic routing technique is applied to find a good feasible bus-route set. Then the greedy feasible solution is improved by the simplex method using column generation technique. The algorithm provides a better near-optimal solution which gives much reductions in the total tour distance and the number of tour buses.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.3
• /
• pp.935-942
• /
• 2000

In this paper we introduce a fast simplification algorithm for terrain height fields to produce a triangulated irregular network, based on the greedy insertion algorithm in [1,4,5]. Our algorithm partitions a terrain height data into rectangular blocks with the same size ad simplifies blocks one by one with the greedy insertion algorithm. Our algorithm references only to the points and the triangles withing each current block for adding a point into the triangulation. Therefore, the algorithm runs faster than the greedy insertion algorithm, which references all input points and triangles in the terrain. Our experiment shows that partitioning method runs from 4 to more than 20 times faster, and it approximates test height fields as accurately as the greedy insertion algorithms. Most greedy insertion algorithms suffer from elongated triangles that usually appear near the boundaries. However, we insert the four corner points into each block to produce the base triangulation of the block before the point addition step begins so that elongated triangles could not appear in th simplified terrain.

• Bulletin of the Korean Mathematical Society
• /
• v.38 no.2
• /
• pp.293-302
• /
• 2001

We improve the greedy algorithm which is one of the general convergence criterion for certain iterative sequence in a given space by building a constructive greedy algorithm on a normed linear space using an arithmetic average of elements. We also show the degree of approximation order is still $Ο(1\sqrt{\n}$) by a bounded linear functional defined on a bounded subset of a normed linear space which offers a good approximation method for neural networks.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.22 no.1
• /
• pp.25-49
• /
• 1997

Printed circuit board (PCB) assembly line configuration is characterized by very long set-up times and high work in process (WIP) inventory level. The scheduling method can significantly reduce the set-up times and WIP inventory level. Greedy sequence dependent scheduling (GSDS) method is proposed based on the current methods. The proposed method is compared with the current method in terms of three performance measures: line throughput, average WIP inventory level, and implementation complexity.

• Proceedings of the KIEE Conference
• /
• 1995.11a
• /
• pp.569-571
• /
• 1995

This paper compares the simulated annealing and the Hopfield neural network method for an optimal routing in a multistage interconnection network(MIN). The MIN provides a multiple number of paths for ATM cells to avoid cell conflict. Exhaustive search always finds the optimal path, but with heavy computation. Although greedy method sets up a path quickly, the path found need not be optimal. The simulated annealing can find an sub optimal path in time comparable with the greedy method.