• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2005.11a
• /
• pp.516-525
• /
• 2005

Case-based reasoning (CBR) has been widely used in various areas due to its convenience and strength in complex problem solving. Generally, in order to obtain successful results from CBR, effective retrieval of useful prior cases for the given problem is essential. However, designing a good matching and retrieval mechanism for CBR systems is still a controversial research issue. Most prior studies have tried to optimize the weights of the features or selection process of appropriate instances. But, these approaches have been performed independently until now. Simultaneous optimization of these components may lead to better performance than in naive models. In particular, there have been few attempts to simultaneously optimize the weight of the features and selection of the instances for CBR. Here we suggest a simultaneous optimization model of these components using a genetic algorithm (GA). We apply it to a customer classification model which utilizes demographic characteristics of customers as inputs to predict their buying behavior for a specific product. Experimental results show that simultaneously optimized CBR may improve the classification accuracy and outperform various optimized models of CBR as well as other classification models including logistic regression, multiple discriminant analysis, artificial neural networks and support vector machines.

• Journal of Power Electronics
• /
• v.15 no.4
• /
• pp.1119-1130
• /
• 2015

This study investigates the multi-objective fuzzy optimization of crowbar resistance for the doubly fed induction generator (DFIG) low-voltage ride-through (LVRT). By integrating the crowbar resistance of the crowbar circuit as a decision variable, a multi-objective model for crowbar resistance value optimization has been established to minimize rotor overcurrent and to simultaneously reduce the DFIG reactive power absorbed from the grid during the process of LVRT. A multi-objective genetic algorithm (MOGA) is applied to solve this optimization problem. In the proposed GA, the value of the crowbar resistance is represented by floating-point numbers in the GA population. The MOGA emphasizes the non-dominated solutions and simultaneously maintains diversity in the non-dominated solutions. A fuzzy-set-theory-based is employed to obtain the best solution. The proposed approach has been evaluated on a 3 MW DFIG LVRT. Simulation results show the effectiveness of the proposed approach for solving the crowbar resistance multi-objective optimization problem in the DFIG LVRT.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.8
• /
• pp.1382-1389
• /
• 2007

This paper presents a new approach for solving the problem of maintenance scheduling of generating units using a binary particle swarm optimization (BPSO). In this paper, we find the optimal solution of the maintenance scheduling of generating units within a specific time horizon using a binary particle swarm optimization algorithm, which is the discrete version of a conventional particle swarm optimization. It is shown that the BPSO method proposed in this paper is effective in obtaining feasible solutions in the maintenance scheduling of generating unit. IEEE reliability test systems(1996) including 32-generators are selected as a sample system for the application of the proposed algorithm. From the result, we can conclude that the BPSO can find the optimal solution of the maintenance scheduling of the generating unit with the desirable degree of accuracy and computation time, compared to other heuristic search algorithm such as genetic algorithms. It is also envisaged that BPSO can be easily implemented for similar optimizations and scheduling problems in power system problems to obtain better solutions and improve convergence performance.

• Journal of Korea Multimedia Society
• /
• v.7 no.6
• /
• pp.841-850
• /
• 2004

This paper proposes a GA and GDM-based method for removing unnecessary rules and generating relevant rules from the fuzzy rules corresponding to several fuzzy partitions. The aim of proposed method is to find a minimum set of fuzzy rules that can correctly classify all the training patterns. When the fine fuzzy partition is used with conventional methods, the number of fuzzy rules has been enormous and the performance of fuzzy inference system became low. This paper presents the application of GA as a means of finding optimal solutions over fuzzy partitions. In each rule, the antecedent part is made up the membership functions of a fuzzy set, and the consequent part is made up of a real number. The membership functions and the number of fuzzy inference rules are tuned by means of the GA, while the real numbers in the consequent parts of the rules are tuned by means of the gradient descent method. It is shown that the proposed method has improved than the performance of conventional method in formulating and solving a combinatorial optimization problem that has two objectives: to maximize the number of correctly classified patterns and to minimize the number of fuzzy rules.

• Journal of KIISE:Computer Systems and Theory
• /
• v.35 no.11
• /
• pp.509-516
• /
• 2008

The Minimum Letter Flips (MLF) model and the Weighted Minimum Letter Flips (WMLF) model are for solving the haplotype assembly problem. But these two models are effective only when the error rate in SNP fragments is low. In this paper, we first establish a new computational model that employs the related genotype information as an improvement of the WMLF model and show its NP-hardness, and then propose an efficient genetic algorithm to solve the haplotype assembly problem. The results of experiments on random data set and a real data set indicate that the introduction of genotype information to the WMLF model is quite effective in improving the reconstruction rate especially when the error rate in SNP fragments is high. And the results also show that genotype information increases the convergence speed of the genetic algorithm.

• International Journal of Control, Automation, and Systems
• /
• v.2 no.2
• /
• pp.247-255
• /
• 2004

The majority of real-world problems encountered by engineers involve simultaneous optimization of competing objectives. In this case instead of single optima, there is a set of alternative trade-offs, generally known as Pareto-optimal solutions. The use of evolutionary algorithms Pareto GA, which was first introduced by Goldberg in 1989, has now become a sort of standard in solving Multiobjective Optimization Problems (MOPs). Though this approach was further developed leading to numerous applications, these applications are based on Pareto ranking and employ the use of the fitness sharing function to maintain diversity. Another scheme for solving MOPs has been presented by J. Nash to solve MOPs originated from Game Theory and Economics. Sefrioui introduced the Nash Genetic Algorithm in 1998. This approach combines genetic algorithms with Nash's idea. Another central achievement of Game Theory is the introduction of an Evolutionary Stable Strategy, introduced by Maynard Smith in 1982. In this paper, we will try to find ESS as a solution of MOPs using our game model based co-evolutionary algorithm. First, we will investigate the validity of our co-evolutionary approach to solve MOPs. That is, we will demonstrate how the evolutionary game can be embodied using co-evolutionary algorithms and also confirm whether it can reach the optimal equilibrium point of a MOP. Second, we will evaluate the effectiveness of our approach, comparing it with other methods through rigorous experiments on several MOPs.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.9 no.1
• /
• pp.221-227
• /
• 2008

A vehicle routing problem with time constraint is one of the important problems in distribution and transportation. The service of a customer must start and finish within a given time interval. Our method is based on an improved operators of genetic algorithm and the objective is to minimize the cost of servicing the set of customers without being tardy or exceeding the capacity or travel time of the vehicles. This research shows that a proposed method based on the improved genetic search can obtain good solutions to vehicle routing problems with time constrained compared with a high degree of efficiency other heuristics. For the computational purpose, we developed a GUI-type computer program according to the proposed method and the computational results show that the proposed method is very effective on a set of standard test problems, and can be potentially useful in solving the vehicle routing problems.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.8
• /
• pp.1415-1421
• /
• 2016

There are considerable concern on the methods for the efficient utilization of the test-resources as increasing of the number of the tests for functionality and performance verification of weapon systems. Furthermore, with an increase in the complexity of the resource assignment the decision support is required. Test resource allocation is basically the same problems as conventional NP-hard FJSP(Flexible Job Shop Problem), therefore empirical test resource allocation method that has been used in many decades is limited in the time performance. Although research has been conducted applying the genetic algorithm to the FJSP, it is limited in the test resource allocation domain in which more than one machine is necessary for a single operation. In this paper, a variable group genetic algorithm is proposed. The algorithm is expected to improve the test plan efficiency by automating and optimizing the existing manual based allocation. The simulation result shows that the algorithm could be applicable to the test plan.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.31 no.2
• /
• pp.141-155
• /
• 2006

In this paper, we address an extended hub-and-spoke transportation network design problem (EHSNP). In the existing hub location problems, the location and number of spokes, and shipments on spokes are given as input data. These may, however, be viewed as the variables according to the areas which they cover. Also, the vehicle routing in each spoke needs to be considered to estimate the network cost more correctly. The EHSNP is a problem of finding the location of hubs and spokes, and pickup/delivery routes from each spoke, while minimizing the total related transportation cost in the network. The EHSNP is an integrated problem that consists of several interrelated sub-problems. To solve EHSNP, we present an approach based on a symbiotic evolutionary algorithm (symbiotic EA), which are known as an efficient tool to solve complex integrated optimization problems. First, we propose a framework of symbiotic EA for EHSNP and its genetic elements suitable for each sub-problem. To analyze the proposed algorithm, the extensive experiments are performed with various test-bed problems. The results show that the proposed algorithm is promising in solving the EHSNP.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.97-108
• /
• 2001

Frauds detection is a difficult problem, requiring huge computer resources and complicated search activities Researchers have struggled with the problem. Even though a fee research approaches have claimed that their solution is much better than others, research community has not found 'the best solution'well fitting every fraud. Because of the evolving nature of the frauds. a novel and self-adapting method should be devised. In this research a new approach is suggested to solving frauds in insurance claims credit card transaction. Based on evolutionary computing approach, the method is itself self-adjusting and evolving enough to generate a new self of decision-makin rules. We believe that this new approach will provide a promising alternative to conventional ones, in terms of computation performance and classification accuracy.