• Proceedings of the IEEK Conference
• /
• 2001.06d
• /
• pp.265-268
• /
• 2001

This paper proposes the strategy of noise injection into inputs in the Kohonen learning algorithm (KKA) to improve the local convergence problem of the KLA. Noise strengths are high in the begin of the learning and gradually lowered as the teaming proceeds. This strategy is a kind of stochastic relaxation (SR) method which is broadly used in the general optimization problems. It is convenient to implement and improves the convergence properties of the KLA with moderately increased computing time compared to the KLA. Experimental results for Gauss-Markov sources and real speech demonstrate that the proposed method can consistently provide better codebooks than the KLA.

• Journal of applied mathematics & informatics
• /
• v.26 no.5_6
• /
• pp.1021-1035
• /
• 2008

In this paper, we analyze an N-design call center with skill-based routing, in which one pool of agents handles two types of calls and another pool of agents handles only one type of calls. The approximate analysis is motivated by a computational complexity that has been observed in the direct stochastic approach and numerical method for finding performance measures. The workforce staffing policy is very important to the successful management of call centers. So the allocation scheduling of the agents can be considered as the optimization problem of the corresponding queueing system to the call center. We use a decomposition algorithm which divides the state space of the queueing system into the subspaces for the approximate analysis of the N-design call center with two different types of agents. We also represent some numerical examples and show the impact of the system parameters on the performance measures.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.32B no.12
• /
• pp.1680-1686
• /
• 1995

GAs, effective stochastic search algorithms based on the model of natural evolution and genetics, have been successfully applied to various optimization problems. When population size is not large, GAs often suffer from the phenomenon of premature convergence in which all chromosomes in the population lose the diversity of genes before they find the optimal solution. In this paper, we propose that a new heuristic that maintains the diversity of genes by adding some chromosomes with random mutation and selective mutation into population during evolution. And population size changes dynamically with supplement of new chromosomes. Experimental results for several test functions show that when population size is rather small and the length of chromosome is not long, this method is effective.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1997.10a
• /
• pp.163-170
• /
• 1997

The present paper describes an investigation into the application of the genetic algorithm(GA) in the optimization of structural design. Stochastic processes generate an initial population of designs and then apply principles of natural selection/survival of the fittest to improve the designs. The five test functions are used to verify the robustness and reliability of GA, and as a numerical example, minimum weight of a cantilever composite laminated beam with a mix of continuous, integer and discrete design variables is obtained by using GA with exterior penalty function method. The design problem has constraints on strength, displacements, and natural frequencies, and is formulated to a multidimensional nonlinear form. From the results, it is found that the GA search technique is very effective at finding the good optimum solution as well as has higher robustness.

• ETRI Journal
• /
• v.37 no.5
• /
• pp.917-921
• /
• 2015

We consider a multiserver system with two classes of customers with preemption, which is a widely used system in the analysis of cognitive radio networks. It is known that the optimal admission control for this system is of threshold type. We express the expected total discounted profit using the total number of customers, thus reducing the stochastic optimization problem with a two-dimensional state space to a problem with a one-dimensional birth-and-death structure. An efficient algorithm is proposed for the calculation of the expected total discounted profit.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.1
• /
• pp.430-437
• /
• 2018

Ultrasonic attenuation is an important parameter in Quantitative Ultrasound and many algorithms have been proposed to improve estimation accuracy and repeatability for multiple independent estimates. In this work, we propose an improved algorithm for estimating ultrasonic attenuation utilizing the optimal frequency compounding technique based on stochastic noise model. We formulate mathematical compounding equations in the AWGN channel model and solve optimization problems to maximize the signal-to-noise ratio for multiple frequency components. Individual estimates are calculated by the reference phantom method which provides very stable results in uniformly attenuating regions. We also propose the guideline to select frequency ranges of reflected RF signals. Simulation results using numerical phantoms show that the proposed optimal frequency compounding method provides improved accuracy while minimizing estimation bias. The estimation variance is reduced by only 16% for the un-compounding case, whereas it is reduced by 68% for the uniformly compounding case. The frequency range corresponding to the half-power for reflected signals also provides robust and efficient estimation performance.

• Investigative Magnetic Resonance Imaging
• /
• v.6 no.2
• /
• pp.137-146
• /
• 2002

Purpose : This paper introduces an improved classification approach which adopts a deterministic relaxation method and an agglomerative clustering technique for the classification of MRI using neural network. The proposed approach can solve the problems of convergency to local optima and computational burden caused by a large number of input patterns when a neural network is used for image classification. Materials and methods : Application of Hopfield neural network has been solving various optimization problems. However, major problem of mapping an image classification problem into a neural network is that network is opt to converge to local optima and its convergency toward the global solution with a standard stochastic relaxation spends much time. Therefore, to avoid local solutions and to achieve fast convergency toward a global optimization, we adopt MFA to a Hopfield network during the classification. MFA replaces the stochastic nature of simulated annealing method with a set of deterministic update rules that act on the average value of the variable. By minimizing averages, it is possible to converge to an equilibrium state considerably faster than standard simulated annealing method. Moreover, the proposed agglomerative clustering algorithm which determines the underlying clusters of the image provides initial input values of Hopfield neural network. Results : The proposed approach which uses agglomerative clustering and deterministic relaxation approach resolves the problem of local optimization and achieves fast convergency toward a global optimization when a neural network is used for MRI classification. Conclusion : In this paper, we introduce a new paradigm to classify MRI using clustering analysis and deterministic relaxation for neural network to improve the classification results.

• The Transactions of the Korean Institute of Electrical Engineers C
• /
• v.51 no.1
• /
• pp.5-14
• /
• 2002

In this paper, we analyze inductive interference in conductive material around 345 kV power transmission line, and evaluate the effects of mitigation wires. Finite element method (FEM) is used to numerically compute induced eddy currents as well as magnetic fields around powder transmission lines. In the analysis model, geometries and electrical properties of various elements such as power transmission line, buried pipe lines, overhead ground wire, and conducting earth are taken into accounts. The calculation shows that mitigation wire reduces fairly good amount of eddy currents in buried pipe line. To find the optimum magnetic field shielding location of mitigation wire, we applied evolution strategy algorithm, a kind of stochastic approach, to the analysis model. Finally, it was shown that we can find more effective shielding effects with optimum location of one mitigation wire than with arbitrary location of multi-mitigation wires around the buried pipe lines.

• IE interfaces
• /
• v.16 no.spc
• /
• pp.76-81
• /
• 2003

BPMS (Business Process Management Systems) is aninformation system that systematically supports designing, administrating, and improving the business processes. It can execute the business processes by assigning tasks to human or computer agents according to the predefined definitions of the processes. In this research we developed a task assignment algorithm that can maximize overall process efficiency under the limitation of agents' capacity. Since BPMS manipulates the formal and predictable business processes, we can analyze the processes using queuing theory to achieve overall process efficiency. We first transform the business processes into queuing network model in which the agents are considered as servers. After that, workloads of agents are calculated as server utilization and we can determine the task assignment policy by balancing the workloads. This will make the workloads of all agents be minimized, and the overall process efficiency is achieved in this way. Another application of the results can be capacity planning of agents in advance and business process optimization in reengineering context. We performed the simulation analysis to validate the results and also show the effectiveness of the algorithm by comparing with well known dispatching policies.

• Journal of Korean Society of Transportation
• /
• v.26 no.5
• /
• pp.175-184
• /
• 2008

Uncertainties are unavoidable in engineering applications. In this paper we propose an alpha reliable multi-variable network design problem under demand uncertainty. In order to decide the optimal capacity enhancement, three performance measures based on 3E(Efficiency, Equity, and Environmental) are considered. The objective is to minimize the total budget required to satisfy alpha reliability constraint of total travel time, equity ratio, and total emission, while considering the route choice behavior of network users. The problem is formulated as the chance-constrained model for application of alpha confidence level and solved as a lexicographic optimization problem to consider the multi-variable. A simulation-based genetic algorithm procedure is developed to solve this complex network design problem(NDP). A simple numerical example ispresented to illustrate the features of the proposed NDP model.