• 정보처리학회논문지B
• /
• 제14B권7호
• /
• pp.523-530
• /
• 2007

본 논문에서는 자동 독순(automatic lipreading)의 인식기로 쓰이는 은닉 마르코프 모델(HMM: hidden Markov model)의 새로운 확률적 최적화 기법을 제안한다. 제안하는 기법은 전역 최적화가 가능한 확률적 기법인 모의 담금질과 지역 최적화 기법을 결합하는 것으로써, 알고리즘의 빠른 수렴과 좋은 해로의 수렴을 가능하게 한다. 제안하는 알고리즘이 전역 최적해로 수렴함을 수학적으로 보인다. 제안하는 기법을 통해 HMM을 학습함으로써 기존의 알고리즘이 지역해만을 찾는 단점을 개선함으로써 향상된 독순 성능을 나타냄을 실험으로 보인다.

• 한국시뮬레이션학회논문지
• /
• 제12권2호
• /
• pp.1-11
• /
• 2003

In this paper, we focus an optimal policy focus optimal class of (s, S) inventory control systems. To this end, we use the perturbation analysis and apply a stochastic optimization algorithm to minimize the average cost over a period. We obtain the gradients of objective function with respect to ordering amount S and reorder point s via a combined perturbation method. This method uses the infinitesimal perturbation analysis and the smoothed perturbation analysis alternatively according to occurrences of ordering event changes. Our simulation results indicate that the optimal estimates of s and S obtained from a stochastic optimization algorithm are quite accurate. We consider that this may be due to the estimated gradients of little noise from the regenerative system simulation, and their effect on search procedure when we apply the stochastic optimization algorithm. The directions for future study stemming from this research pertain to extension to the more general inventory system with regard to demand distribution, backlogging policy, lead time, and review period. Another directions involves the efficiency of stochastic optimization algorithm related to searching procedure for an improving point of (s, S).

• 전기학회논문지
• /
• 제59권9호
• /
• pp.1600-1604
• /
• 2010

This paper presents the optimum design of the brushless motor to maximize the output power per weight considering the design parameter tolerance. The optimization is proceeded by commercial software that is adopted the scatter-search algorithm and the characteristic analysis is conducted by FEM. The stochastic optimum design results are compared with those of the deterministic optimization method. We can verify that the results of the stochastic optimization is superior than that of deterministic optimization.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1997년도 하계학술대회 논문집 B
• /
• pp.699-701
• /
• 1997

The simulated annealing(SA) algorithm is a stochastic strategy for search of the ground state and a powerful tool for optimization, based on the annealing process used for the crystallization in physical systems. It's main disadvantage is the long convergence time. Therefore, this paper proposes a stochastic algorithm combined with conventional deterministic optimization method to reduce the computation time, which is called SDS(Stochastic-Deterministic-Stochastic) method.

• 전자공학회논문지B
• /
• 제31B권4호
• /
• pp.145-154
• /
• 1994

This paper proposes an efficient method for improving the training performance of the neural network by using a hybrid of a stochastic approximation and a backpropagation algorithm. The proposed method improves the performance of the training by appliying a global optimization method which is a hybrid of a stochastic approximation and a backpropagation algorithm. The approximate initial point for a stochastic approximation and a backpropagation algorihtm. The approximate initial point for fast global optimization is estimated first by applying the stochastic approximation, and then the backpropagation algorithm, which is the fast gradient descent method, is applied for a high speed global optimization. And further speed-up of training is made possible by adjusting the training parameters of each of the output and the hidden layer adaptively to the standard deviation of the neuron output of each layer. The proposed method has been applied to the parity checking and the pattern classification, and the simulation results show that the performance of the proposed method is superior to that of the backpropagation, the Baba's MROM, and the Sun's method with randomized initial point settings. The results of adaptive adjusting of the training parameters show that the proposed method further improves the convergence speed about 20% in training.

• 대한수학회보
• /
• 제55권4호
• /
• pp.1303-1315
• /
• 2018

For principal component analysis (PCA) to efficiently analyze large scale matrices, it is crucial to find a few singular vectors in cheaper computational cost and under lower memory requirement. To compute those in a fast and robust way, we propose a new stochastic method. Especially, we adopt the stochastic variance reduced gradient (SVRG) method [11] to avoid asymptotically slow convergence in stochastic gradient descent methods. For that purpose, we reformulate the PCA problem as a unconstrained optimization problem using a quadratic penalty. In general, increasing the penalty parameter to infinity is needed for the equivalence of the two problems. However, in this case, exact penalization is guaranteed by applying the analysis in [24]. We establish the convergence rate of the proposed method to a stationary point and numerical experiments illustrate the validity and efficiency of the proposed method.

• 한국시뮬레이션학회논문지
• /
• 제22권4호
• /
• pp.21-28
• /
• 2013

일반적으로 최적화 문제에서 군사 시뮬레이션과 같이 결과가 확률적으로 나타나는 경우를 계산할 때에는 문제를 모델링 하여 일반적인 최적화 기법을 적용하는 것에 어려움이 있다. 본 논문에서는 이러한 군사 시뮬레이션의 특징을 반영하는 복잡한 반응표면을 가진 확률적 평가 함수를 정의하였다. 그리고 이러한 확률적 시뮬레이션에 대해 기존의 PSO법이 가진 약점을 보완하는 기법을 제안하였다. 제안한 기법을 이용해 평가 함수에 대한 최적화를 시행하였으며 최적화의 속도와 정확도에 영향을 미치는 계산 조건들의 상호작용을 분석하였다. 이를 통해 본 논문에서 제안한 확률적 시뮬레이션의 최적화 전략을 제시하였다.

• 대한전기학회논문지:전기기기및에너지변환시스템부문B
• /
• 제49권4호
• /
• pp.219-225
• /
• 2000

This paper presents the shape optimization considering the uncertainty of design variable to find robust optimal solution that has insensitive performance to its change of design variable. Stochastic finite element method (SFEM) is used to treat input data as stochastic variables. It is method that the potential values are series form for the expectation and small variation. Using correlation function of their variables, the statistics of output obtained form the input data distributed. From this, design considering uncertainty of design variables.

• 한국시뮬레이션학회논문지
• /
• 제3권2호
• /
• pp.57-67
• /
• 1994

For many practical and industrial optimization problems where some or all of the system components are stochastic, the objective functions cannot be represented analytically. Therefore, modeling by computer simulation is one of the most effective means of studying such complex systems. In this paper, with discussion of simulation optimization techniques, a case study in machining process for application of simulation optimization is presented. Most of optimization techniques can be classified as single-or multiple-response techniques. The optimization of single-response category, these strategies are gradient based search methods, stochastic approximate method, response surface method, and heuristic search methods. In the multiple-response category, there are basically five distinct strategies for treating the responses and finding the optimum solution. These strategies are graphical method, direct search method, constrained optimization, unconstrained optimization, and goal programming methods. The choice of the procedure to employ in simulation optimization depends on the analyst and the problem to be solved.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제19권4호
• /
• pp.417-428
• /
• 2015

Although, in general, the random fluctuation of interest rates gives a limited impact on portfolio optimization, their stochastic nature may exert a significant influence on the process of selecting the proportions of various assets to be held in a given portfolio when the stochastic volatility of risky assets is considered. The stochastic volatility covers a variety of known models to fit in with diverse economic environments. In this paper, an optimal strategy for portfolio selection as well as the smoothness properties of the relevant value function are studied with the dynamic programming method under a market model of both stochastic volatility and stochastic interest rates.