http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
유전적 프로그래밍을 이용한 응답면의 모델링 I : 방향도함수 기반의 Smoothering 기법
연윤석,이욱,Yeun, Yun-Seog,Rhee, Wook 한국정보기술응용학회 2001 정보기술응용연구 Vol.3 No.3
This paper introduces the genetic programming algorithm(GP), which can approximate highly nonlinear functions, as a tool for the modeling of response surfaces. When the response surfaces is approximated, the very small or minimal teaming set should be used, and thus it is almost certain that GP trees will show overfilling that must be avoided at all costs. We present a novel method, calledDDBS(DirectionalDerivative-Based Smoothering), which very effectively eliminates the unwanted behaviors of GP trees such as large peaks, oscillations, and also overfitting. Four illustrative numerical examples are given to demonstrate the performance of the genetic programming algorithm that adopts DDBS.
Managing Approximation in Collaborative Optimization
Young-Soon Yang(양영순),Beom-Seon Jang(장범선),Yun-Seog Yeun(연윤석) 대한기계학회 2001 대한기계학회 춘추학술대회 Vol.2001 No.10
This paper describes the use of approximation in Collaborative Optimization (CO) method, one of the Multidisciplinary Design Optimization (MDO) techniques. The approximation is used to model the result of a disciplinary design, optimal discrepancy function value, as a function of the interdisciplinary target variables passed from system level to the discipline. The optimal discrepancy function value is used to examine the interdisciplinary compatibility constraint (discrepancy function =0) during the system level optimization. However, the peculiar shape of the compatibility constraint makes it difficult to exploit well-developed conventional approximation methods. This paper introduces the combination of neural network classification and kriging to resolve this problem. In addition, for the purpose of enhancing the accuracy of the approximation, the approximation is continuously updated using the information obtained from the system level optimization. This iterative process is continued until acceptable convergence is achieved.
연윤석 대진대학교 생산기술연구소 1998 생산기술연구소 논문집 Vol.1 No.-
This paper deals with smooth fitting of data corrupt by noise. Most research efforts have been concentrated on employing the smoothness penalty function with the stimation of its optimal parameter in order to avoid the "overfitting and underfitting "dilemma in noisy data fitting problems. My approach, called DBSF(Differentiation-Based Smooth Fitting), is different from the above-mentioned method. The amin idea is that optimal functions approximately estimating the derivative of noisy data are generated first using genetic programming, and then their integral values are evaluated and used to recover original data. To show the effectiveness of this approach, DBSP is demonstrated by presenting two illustrative examples and the application of estimating the principal dimensions of bulk carog ships in the conceptual design stage.