• Journal of Korea Society of Industrial Information Systems
• /
• v.2 no.1
• /
• pp.1-17
• /
• 1997

Genetic algorithms have proved to be a versatile and effectvie approach for solving optimization problems. Nevertheless, there are many situations that the genetic algorithm does not perform particularly well, and so various methods of hybridization have been proposed. Thus, this paper develop a hybrid method and a precision search method around optimum in the gentic algorithm and the conventional optimization techniques in finding global or near optimum.

• Journal of Korea Society of Industrial Information Systems
• /
• v.19 no.4
• /
• pp.87-104
• /
• 2014

This paper proposes regionally centralized multistage reverse logistics (cmRL) networks and regionally decentralized multistage reverse logistics (dmRL) networks. cmRL considers whole area of RL network, while dmRL regionally distributed area of RL network. Each type is formulated by the mixed integer programming (MIP) models, which are realized in genetic algorithm (GA) approach. Two types of numerical experiments using RL network are presented and various measures of performance are used for comparing the efficiency of the cmRL and the dmRL. Finally, it is proved that the performance of the cmRL is superior to that of the dmRL.

• Journal of Korea Society of Industrial Information Systems
• /
• v.27 no.1
• /
• pp.37-48
• /
• 2022

In this study, a hybrid meta-heuristic approach is proposed for solving engineering design optimization problems. Various approaches in many literatures have been proposed to solve engineering optimization problems with various types of decision variables and complex constraints. Unfortunately, however, their efficiencies for locating optimal solution do not be highly improved. Therefore, we propose a hybrid meta-heuristic approach for improving their weaknesses. the proposed GA-VNS-HC approach is combining genetic algorithm (GA) for global search with variable neighborhood search (VNS) and hill climbing (HC) for local search. In case study, various types of engineering design optimization problems are used for proving the efficiency of the proposed GA-VNS-HC approach

• Journal of Korea Society of Industrial Information Systems
• /
• v.1 no.1
• /
• pp.63-85
• /
• 1996

This paper describes the applications of genetic algorithm to nonlinear constrained optimization problems. Genetic algorithms are combinatorial in nature, and therefore are computationally suitable for treating continuous and idstrete integer design variables. For several problems , the conventional genetic algorithms are ill-defined , which comes from the application of penalty function , encoding and decoding methods, fitness scaling, and premature convergence of solution. Thus, we develope a hybrid genetic algorithm to resolve these problems and present two examples to demonstrate the effectiveness of the methodology developed in this paper.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.11 no.6
• /
• pp.939-945
• /
• 2000

A cylindrical object buried in the lossy half space is reconstructed from the measured scattered fields above the lossy half space. The position, the size and the medium parameters i.e. relative dielectric constants and conductivity of the buried object as well as the medium parameters of the background lossy half space are obtained from the scattered fields by using the iterative inversion method and the optimization hybrid algorithm combining the genetic algorithm and the Levenberg-Marquardt algorithm. Illposedness of the inversion due to the measurement errors in the scattered fields are regularized by filtering out the evanescent modes in the spatial frequency spectrum domain.

• Journal of the Society of Naval Architects of Korea
• /
• v.31 no.4
• /
• pp.147-156
• /
• 1994

Though optimization method had been used for long time for the optimal design of ship structure, design variables in the most cases were assumed to be continuous real values or it was not easy to solve the mixed integer optimum design problems using the conventional optimization methods. Thus, it was often tried to use various initial starting points to locate the best optimum paint and to use special method such as branch and bound method to handle the discrete design variables in the optimization problems. Sometimes it had succeed, but the essential problems for dealing with the local optimum and discrete design variables was left unsolved. Hence, in this paper, Genetic Algorithms adopting the biological evolution process is applied to the ship structural design problem where the integer values for the number of stiffen design variables or the discrete values for the plate thickness variables would be more preferable in order to find out their effects on the final optimum design. Through the numerical result comparisons, it was found that Genetic Algorithm could always yield the global optimum for the discrete and mixed integer structural optimization problem cases even though it takes more time than other methods.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.25 no.4
• /
• pp.425-433
• /
• 2022

This paper studies the method of delivering munitions using both drones and UGVs that are developing along with the 4th Industrial Revolution. While drones are more mobile than UGVs, their loading capacity is small, and UGVs have relatively less mobility than drones, but their loading capacity is better. Therefore, by simultaneously operating these two delivery means, each other's shortcomings may be compensated. In addition, on actual battlefields, time constraints are an important factor in delivering munitions. Therefore, assuming an actual battlefield environment with a time limit, we establish delivery routes that minimize delivery time by operating both drones and UGVs with different capacities and speeds. If the delivery is not completed within the time limit, penalties are imposed. We devised the hybrid genetic algorithm to find solutions to the proposed model, and as results of the experiment, we showed the algorithm we presented solved the actual size problems in a short time.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.23 no.3
• /
• pp.1-15
• /
• 1998

Genetic algorithms(GAs) are suited for solving structural design problems, since they handle the design variables efficiently. This ability of GAs considers then as a good choice for optimization problems. Nevertheless, there are many situations that the conventional genetic algorithms do not perform particularly well, and so various methods of hybridization have been proposed. Thus. this paper develops a hybrid genetic algorithm(HGA) to incorporate a local convergence method and precision search method around optimum in the genetic algorithms. In case study. it is showed that HGA is able consistently to provide efficient, fine quality solutions and provide a significant capability for solving structural design problems.

• Proceedings of the KSME Conference
• /
• 2007.05b
• /
• pp.2476-2481
• /
• 2007

An inverse radiation analysis is presented for the estimation of the wall emissivities for an absorbing, emitting, and scattering media with diffusely emitting and reflecting opaque boundaries. In this study, a repulsive particle swarm optimization(RPSO) algorithm which is a relatively recent heuristic search method is proposed as an effective method for improving the search efficiency for unknown parameters. To verify the performance of the proposed RPSO algorithm, it is compared with a basic particle swarm optimization(PSO) algorithm and a hybrid genetic algorithm(HGA) for the inverse radiation problem with estimating the wall emissivities in a two-dimensional irregular medium when the measured temperatures are given at only four data positions. A finite-volume method is applied to solve the radiative transfer equation of a direct problem to obtain measured temperatures.

• The Korean Journal of Applied Statistics
• /
• v.34 no.2
• /
• pp.153-166
• /
• 2021

A main goal of pharmacogenomics studies is to predict individual's drug responsiveness based on high dimensional genetic variables. Due to a large number of variables, feature selection is required in order to reduce the number of variables. The selected features are used to construct a predictive model using machine learning algorithms. In the present study, we applied several hybrid feature selection methods such as combinations of logistic regression, ReliefF, TurF, random forest, and LASSO to a next generation sequencing data set of 400 epilepsy patients. We then applied the selected features to machine learning methods including random forest, gradient boosting, and support vector machine as well as a stacking ensemble method. Our results showed that the stacking model with a hybrid feature selection of random forest and ReliefF performs better than with other combinations of approaches. Based on a 5-fold cross validation partition, the mean test accuracy value of the best model was 0.727 and the mean test AUC value of the best model was 0.761. It also appeared that the stacking models outperform than single machine learning predictive models when using the same selected features.