• 통합자연과학논문집
• /
• 제2권2호
• /
• pp.55-58
• /
• 2009

A ligand-receptor docking program is an indispensible tool in modern pharmaceutical design. An accurate prediction of small molecular docking pose to a receptor is essential in drug design as well as molecular recognition. An effective docking program requires the ability to locate a correct binding pose in a surprisingly complex conformational space. However, there is an inherent difficulty to predict correct binding pose. The odds are more demanding than finding a needle in a haystack. This mainly comes from the flexibility of both ligand and receptor. Because the searching space to consider is so vast, receptor rigidity has been often applied in docking programs. Even nowadays the receptor may not be considered to be fully flexible although there have been some progress in search algorithm. Improving the efficiency of searching algorithm is still in great demand to explore other applications areas with inherently flexible ligand and/or receptor. In addition to classical search algorithms such as molecular dynamics, Monte Carlo, genetic algorithm and simulated annealing, rather recent algorithms such as tabu search, stochastic tunneling, particle swarm optimizations were also found to be effective. A good search algorithm would require a good balance between exploration and exploitation. It would be a good strategy to combine algorithms already developed. This composite algorithms can be more effective than an individual search algorithms.

• Steel and Composite Structures
• /
• 제18권2호
• /
• pp.289-303
• /
• 2015

This study presents a strategy so-called Subspace Search Mechanism (SSM) for reducing the computational time for convergence of population based metaheusristic algorithms. The selected metaheuristic for this study is the Cuckoo Search algorithm (CS) dealing with size optimization of trusses. The complexity of structural optimization problems can be partially due to the presence of high-dimensional design variables. SSM approach aims to reduce dimension of the problem. Design variables are categorized to predefined groups (subspaces). SSM focuses on the multiple use of the metaheuristic at hand for each subspace. Optimizer updates the design variables for each subspace independently. Updating rules require candidate designs evaluation. Each candidate design is the assemblage of responsible set of design variables that define the subspace of interest. SSM is incorporated to the Cuckoo Search algorithm for size optimizing of three small, moderate and large space trusses. Optimization results indicate that SSM enables the CS to work with less number of population (42%), as a result reducing the time of convergence, in exchange for some accuracy (1.5%). It is shown that the loss of accuracy can be lessened with increasing the order of complexity. This suggests its applicability to other algorithms and other complex finite element-based engineering design problems.

• 한국콘텐츠학회논문지
• /
• 제14권11호
• /
• pp.541-548
• /
• 2014

본 논문에서는 유전알고리즘의 유전자형-표현형 기법을 적용한 수정된 이진 입자군집최적화의 두 번째 버전을 소개한다. 입자군집최적화는 해를 탐색해 나가는 과정에서 주변의 우수한 해의 위치와 자신의 위치차이 정보를 이용한다. 이러한 위치 차이를 구하는데 있어서 첫 번째 버전의 수정된 이진 입자군집최적화는 표현형을 사용한 반면에 제안하는 버전은 유전자형을 사용한다. 이진 정보만을 제공하는 표현형에 비해 연속 공간 전체를 탐색공간으로 제공하는 유전자형 정보를 사용하여 해 공간을 보다 넓은 공간으로 표시할 수 있다. 벤치마크 함수인 10개의 De Jong 함수에 실험한 결과, 두 번째 버전은 탐색 공간이 넓고 지역 최적해가 많은 함수에서 첫 번째 버전에 보다 우수한 결과를 얻었다.

• 한국지능시스템학회논문지
• /
• 제13권4호
• /
• pp.381-385
• /
• 2003

본 논문에서는 다면체의 인식을 위하여 사용되는 여러-방향-보기 방법 (multiple-view approach)에서, ART-1 신경망을 이용하여 모델베이스의 탐색공간 크기를 줄이기 위한 방법을 제안한다. 이 방법에서 모델베이스는 물체를 둘러싸고 있는 보기 구체의 미리 정해진 시점에서 관측된 2차원 투영체에서 추출된 특징들로 구성된다.

• Structural Monitoring and Maintenance
• /
• 제1권4호
• /
• pp.427-449
• /
• 2014

A hybrid optimization method for the identification of state-space models is presented in this study. Hybridization is succeeded by combining the advantages of deterministic and stochastic algorithms in a superior scheme that promises faster convergence rate and reliability in the search for the global optimum. The proposed hybrid algorithm is developed by replacing the original stochastic mutation operator of Evolution Strategies (ES) by the Levenberg-Marquardt (LM) quasi-Newton algorithm. This substitution results in a scheme where the entire population cloud is involved in the search for the global optimum, while single individuals are involved in the local search, undertaken by the LM method. The novel hybrid identification framework is assessed through the Monte Carlo analysis of a simulated system and an experimental case study on a shear frame structure. Comparisons to subspace identification, as well as to conventional, self-adaptive ES provide significant indication of superior performance.

• Journal of the Optical Society of Korea
• /
• 제19권1호
• /
• pp.45-54
• /
• 2015

In this study, a novel parallel wavefront correction system architecture is proposed, and a model-based tabu search (MBTS) algorithm is introduced for this new system to compensate wavefront aberration caused by atmospheric turbulence in a free-space optical (FSO) communication system. The algorithm flowchart is presented, and a simple hypothetical design for the parallel correction system with multiple adaptive optical (AO) subsystems is given. The simulated performance of MBTS for an AO-FSO system is analyzed. The results indicate that the proposed algorithm offers better performance in wavefront aberration compensation, coupling efficiency, and convergence speed than a stochastic parallel gradient descent (SPGD) algorithm.

• 전자공학회논문지C
• /
• 제35C권2호
• /
• pp.48-55
• /
• 1998

Structural learning methods of MLP classifiers for a given application using genetic algorithms have been studied. In the methods, however, the search space for an optimal structure is increased exponentially for the physical application of high diemension-multi calss. In this paperwe propose a method of MLP classifiers using species genetic algorithm(SGA), a modified GA. In SGA, total search space is divided into several subspaces according to the number of hidden units. Each of the subdivided spaces is called "species". We eliminate low promising species from the evoluationary process in order to reduce the search space. experimental results show that the proposed method is more efficient than the conventional genetic algorithm methods in the aspect of the misclassification ratio, the learning rate, and the structure.structure.

• 한국정보과학회논문지:데이타베이스
• /
• 제31권2호
• /
• pp.83-90
• /
• 2004

본 논문에서는 순환적인 검색공간과 장애물이 존재하는 검색공간에서 객체들을 클러스터링할 때 고려해야하는 CCD(Clustering with Circular Distance) 문제와 COD(Clustering with Obstructed Distance) 문제를 연구하였다. 그리고 다차원 검색공간에서 삽입이나 삭제가 빈번히 발생하는 객체들을 효율적으로 클러스터링하기 위한 새로운 클러스터링 알고리즘을 제안하였다. 제안한 클러스터링 알고리즘에는 CCD 및 COD 문제를 해결하기 위한 거리 함수가 정의된다. 그리고 최소의 연산 시간으로 높은 공간 국부성을 갖는 클러스터들을 생성하기 위한 클러스터링 방법이 포함된다.

• 정보처리학회논문지D
• /
• 제8D권2호
• /
• pp.145-153
• /
• 2001

피라미드 기법은 n-차원 공간 데이터를 1차원 데이터로 변환하여 $B^+$-트리로 표현하고, n-차원 데이터 공간에서 하이퍼큐브 영역질의 처리로 발생하는 "차원의 저주현상"에 영향을 받지 않게 검색 시간 문제를 해결하고 있다. 구형 피라미드 기법은 피라미드 기법의 공간 분할 전략을 응용하여 유사도 검색에 적합하도록 구 영역질의 방법을 사용하고 검색 성능을 개선하고 있다. 그러나 두 방법은 데이터 크기와 차원 변화에 따른 검색 성능이 100만건 이상과 16차원 이상일 때 현저하게 저하하는 현상을 보이고 있다. 이 논문에서는 멀티미디어 데이터와 같은 고차원 데이터의 검색 성능을 향상시키기 위한 새로운 인덱스 구조로 PdR-트리를 제안한다. 모의 데이터와 실제 데이터를 이용하여 실험한 결과, PdR-트리가 피라미드 기법과 구형 피라미드 기법보다 검색 성능이 향상되었음을 보이고 있다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.86-89
• /
• 2003

To model a numerical problem space under the limitation of available data, we need to extract sparse but key points from the space and to efficiently approximate the space with them. This study proposes a sampling method based on the search process of genetic algorithm and a space modeling method based on least-squares approximation using the summation of Gaussian functions. We conducted simulations to evaluate them for several kinds of problem spaces: DeJong's, Schaffer's, and our original one. We then compared the performance between our sampling method and sampling at regular intervals and that between our modeling method and modeling using a polynomial. The results showed that the error between a problem space and its model was the smallest for the combination of our sampling and modeling methods for many problem spaces when the number of samples was considerably small.