• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.27 no.12
• /
• pp.2047-2054
• /
• 2003

In this paper, an approach of automatically finding and modifying the most appropriate orthogonal array (GO) is suggested and applied to a new structural optimization procedure with two steps. GO is motivated by the situation where finding a proper orthogonal array from the tables in the literature is difficult or impossible. Now the Taguchi method is made available for various numbers of variables and levels. In the two-step structural optimization, the Taguchi method equipped with GO and a shape optimization using the finite differencing method is consecutively applied. The existence or non-existence of an element can be taken as a factor level and this feature is utilized finding the best topology from a set of potential topologies suggested from the user's expertise. This greatly enhances applicability and one can expect a better result than the case in which each step is applied independently because these steps are complementary each other.

• Journal of Computing Science and Engineering
• /
• v.11 no.4
• /
• pp.121-129
• /
• 2017

Feature selection has been widely established as an efficient technique for microarray data analysis. Feature selection aims to search for the most important feature/gene subset of a given dataset according to its relevance to the current target. Unsupervised feature selection is considered to be challenging due to the lack of label information. In this paper, we propose a novel method for unsupervised feature selection, which incorporates embedded learning and $l_{2,1}-norm$ sparse regression into a framework to select genes in microarray data analysis. Local tangent space alignment is applied during embedded learning to preserve the local data structure. The $l_{2,1}-norm$ sparse regression acts as a constraint to aid in learning the gene weights correlatively, by which the proposed method optimizes for selecting the informative genes which better capture the interesting natural classes of samples. We provide an effective algorithm to solve the optimization problem in our method. Finally, to validate the efficacy of the proposed method, we evaluate the proposed method on real microarray gene expression datasets. The experimental results demonstrate that the proposed method obtains quite promising performance.

• International Journal of CAD/CAM
• /
• v.6 no.1
• /
• pp.89-97
• /
• 2006

This paper presents a novel approach for non-iterative surface smoothing with feature preservation on arbitrary meshes. Laplacian operator is performed in a global way over the mesh. The surface smoothing is formulated as a quadratic optimization problem, which is easily solved by a sparse linear system. The cost function to be optimized penalizes deviations from the global Laplacian operator while maintaining the overall shape of the original mesh. The features of the original mesh can be preserved by adding feature constraints and barycenter constraints in the system. Our approach is simple and fast, and does not cause surface shrinkage and distortion. Many experimental results are presented to show the applicability and flexibility of the approach.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.11
• /
• pp.5592-5609
• /
• 2017

JPEG steganography detection is an active research topic in the field of information hiding due to the wide use of JPEG image in social network, image-sharing websites, and Internet communication, etc. In this paper, a new steganalysis method for content-adaptive JPEG steganography is proposed by integrating the evolutionary feature selection and classifier ensemble selection. First, the whole framework of the proposed steganalysis method is presented and then the characteristic of the proposed method is analyzed. Second, the feature selection method based on genetic algorithm is given and the implement process is described in detail. Third, the method of classifier ensemble selection is proposed based on Pareto evolutionary optimization. The experimental results indicate the proposed steganalysis method can achieve a competitive detection performance by compared with the state-of-the-art steganalysis methods when used for the detection of the latest content-adaptive JPEG steganography algorithms.

• Journal of information and communication convergence engineering
• /
• v.3 no.4
• /
• pp.209-212
• /
• 2005

This study proposes hybrid case-based reasoning and genetic algorithms model for customer classification. In this study, vertical and horizontal dimensions of the research data are reduced through integrated feature and instance selection process using genetic algorithms. We applied the proposed model to customer classification model which utilizes customers' demographic characteristics as inputs to predict their buying behavior for the specific product. Experimental results show that the proposed model may improve the classification accuracy and outperform various optimization models of typical CBR system.

• Proceedings of the Korean Society of Soil and Groundwater Environment Conference
• /
• 2002.09a
• /
• pp.24-27
• /
• 2002

The optimal designs of groundwater problems of contaminant containment and cleanup using linear programming and genetic algorithm are provided. In the containment problem, genetic algorithm shows the superior feature to linear programming. In cleanup problem, genetic algorithm makes reasonable optimal design. Un this study, it is demonstrated through numerical experiments that genetic algorithm can be applied to remedial designs of groundwater problems.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.708-711
• /
• 1997

This paper proposes the hybrid algorithm for the optimization of the structure and parameters of the fuzzy neural networks by genetic algorithms (GA) to improve the behaviour and the design of fuzzy neural networks. Fuzzy neural networks have a distinguishing feature in that they can possess the advantage of both neural networks and fuzzy systems. In this way, we can bring the low-level learning and computational power of neural networks into fuzzy systems and also high-level, human like IF-THEN rule thinking and reasoning of fuzzy systems into neural networks. As a result, there are many research works concerning the optimization of the structure and parameters of fuzzy neural networks. In this paper, we propose the hybrid algorithm that can optimize both the structure and parameters of fuzzy neural networks. Numerical example is provided to show the advantages of the proposed method.

• Journal of Magnetics
• /
• v.19 no.1
• /
• pp.78-83
• /
• 2014

This paper proposes an efficient reliability-based optimization method for designing a superconducting magnetic energy system in presence of uncertainty. To evaluate the probability of failure of constraints, samplingbased reliability analysis method is employed, where Monte Carlo simulation is incorporated into dynamic Kriging models. Its main feature is to drastically reduce the numbers of iterative designs and computer simulations during the optimization process without sacrificing the accuracy of reliability analysis. Through comparison with existing methods, the validity of the proposed method is examined with the TEAM Workshop Problem 22.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.14 no.6
• /
• pp.90-97
• /
• 2015

Optimization of multi-quality characteristics in injection molded parts is very important, but it is sometimes difficult for part/mold designers. The objective of this study is to develop a practical design methodology for optimizing two-quality characteristics of injection molded parts. To attain this end, we developed a new design-range reduction algorithm based on Taguchi's orthogonal arrays for two characteristics. Then, the algorithm was integrated with commercial injection-molding simulation tools. A feature of the proposed methodology is that it allows field-designers unfamiliar with general optimization methods to be able to apply the methodology to their design problems with ease. Finally, we have applied the proposed design methodology to optimization of weldlines and deflections in an actual bezel model. The results show the usefulness of this methodology.

• Advances in Computational Design
• /
• v.1 no.2
• /
• pp.119-138
• /
• 2016

In this article, hybridization of enhanced colliding bodies optimization (ECBO) with upper bound strategy (UBS) that is called UECBO is proposed for optimum design of truss structures with frequency constraints. The distinct feature of the proposed algorithm is that it requires less computational time while preserving the good accuracy of the ECBO. Four truss structures with frequency limitations selected from the literature are studied to verify the viability of the algorithm. This type of problems is highly non-linear and non-convex. The numerical results show the successful performance of the UECBO algorithm in comparison to the CBO, ECBO and some other metaheuristic optimization methods.