• Journal of Intelligence and Information Systems
• /
• v.4 no.1
• /
• pp.133-147
• /
• 1998

Recently, many researchers have been involved in finding deterministic equations which can accurately predict future event, based on chaotic theory, or fractal theory. The theory says that some events which seem very random but internally deterministic can be accurately predicted by fractal equations. In contrast to the conventional methods, such as AR model, MA, model, or ARIMA model, the fractal equation attempts to discover a deterministic order inherent in time series data set. In discovering deterministic order, researchers have found that neural networks are much more effective than the conventional statistical models. Even though prediction accuracy of the network can be different depending on the topological structure and modification of the algorithms, many researchers asserted that the neural network systems outperforms other systems, because of non-linear behaviour of the network models, mechanisms of massive parallel processing, generalization capability based on adaptive learning. However, recent survey shows that prediction accuracy of the forecasting models can be determined by the model structure and data structures. In the experiments based on actual economic data sets, it was found that the prediction accuracy of the neural network model is similar to the performance level of the conventional forecasting model. Especially, for the data set which is deterministically chaotic, the AR model, a conventional statistical model, was not significantly different from the MLP model, a neural network model. This result shows that the forecasting model. This result shows that the forecasting model a, pp.opriate to a prediction task should be selected based on characteristics of the time series data set. Analysis of the characteristics of the data set was performed by fractal analysis, measurement of Hurst index, and measurement of Lyapunov exponents. As a conclusion, a significant difference was not found in forecasting future events for the time series data which is deterministically chaotic, between a conventional forecasting model and a typical neural network model.

• Journal of applied mathematics & informatics
• /
• v.27 no.3_4
• /
• pp.621-638
• /
• 2009

The linear system Ax = b will have (i) no solution, (ii) only one non-trivial (trivial) solution, or (iii) infinity of solutions. Our focus will be on cases (ii) and (iii). The mathematical models of many real-world problems give rise to (a) ill-conditioned linear systems, (b) singular linear systems (A is singular with all its linearly independent rows are sufficiently linearly independent), or (c) ill-conditioned singular linear systems (A is singular with some or all of its strictly linearly independent rows are near-linearly dependent). This article highlights the scope and need of a randomized algorithm for ill-conditioned/singular systems when a reasonably narrow domain of a solution vector is specified. Further, it stresses that with the increasing computing power, the importance of randomized algorithms is also increasing. It also points out that, for many optimization linear/nonlinear problems, randomized algorithms are increasingly dominating the deterministic approaches and, for some problems such as the traveling salesman problem, randomized algorithms are the only alternatives.

• 전기의세계
• /
• v.22 no.2
• /
• pp.28-32
• /
• 1973

System modeling is applied in developing a probabilistic linear estimator for the load of an electric power system for the purpose of short term load forecasting. The model assumer that the load in given by the suns of a periodic discrete time serier with a period of 24 hour and a residual term such that the output of a discrete time dynamical linear system driven by a white random process and a deterministic input. And also we have established the main forecasting algorithms, which are essemtally the Kalman filter-predictor equations.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.238-240
• /
• 2002

진화 가능한 하드웨어의 개발은 유전자 알고리즘의 새로운 가능성을 열어주었고 이에 적합한 다양한 방법이 제시되어 왔다. 하지만 일반적인 유전자 알고리즘으로는 Genetic drift가 생기거나 지역해에 빠지는 등 한계가 있기 때문에 이를 해결하기 위한 방안으로 종분화 알고리즘이 도입되고 있다. 현재까지 다양한 종분화 알고리즘이 소개되었는데 이들은 이전의 알고리즘과 비교하였을 때 높은 다양성을 유지하면서 더 좋은 해를 찾아낸다. 이 논문에서는 진화 하드웨어상에서 이러한 종분화 알고리즘들의 장단점 및 특징을 여러 비교기준을 통해 제시한다. 실험결과 Deterministic Crowding과 Struggle GA가 가장 좋은 성능을 나타내었다.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.13 no.21
• /
• pp.51-60
• /
• 1990

In this paper we deal with the miniaml cost network flow problem with uncertain arc capacity constraints. When the arc capacities are fuzzy with linear L-R type membership function, using parametric programming procedure, we reduced it to the deterministic minimal cost network flow problem which can be solved by various typical network flow algorithms. A modified Algorithm using the Out-of-kilter algorithm is developed.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.9 no.2
• /
• pp.15-22
• /
• 1984

A deterministic multiproduct, facility-in series multiperiod production planning model is analyzed, where each period demand for the product of a facility appear in a fixed proportion of that for the product of the immediately following facility. The model considers concave production and inventory costs, which can depend upon the production in different facilities. No backlogging is allowed. It is shown that the model is represented via a single source network, which facilitates development of efficient dynamic programming algorithms for computing the optimal production schedule.

• Journal of the Korea Society of Computer and Information
• /
• v.22 no.9
• /
• pp.81-89
• /
• 2017

The CRP(Container Relocation Problem) algorithms pursuing efficient container relocation of wharf container terminal can not be deterministic because of the large number of layout cases. Therefore, the CRP algorithms should adopt trial and error intuition and experimental heuristic techniques. And because the heuristic can not be best for all individual cases, it is necessary to find metrics which show excellent on average. In this study, we analyze GLAH(Greedy Look-ahead Heuristic) algorithm which is one of the recent researches in detail, and propose a heuristic metrics HOB(sum of the height differences between a badly placed container and the containers prohibited by the badly placed container) to improve the algorithm. The experimental results show that the improved algorithm, GLAH', exerts a stable performance increment of up to 3.8% in our test data, and as the layout size grows, the performance increment gap increases.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.3 no.1
• /
• pp.115-125
• /
• 1999

Simple Genetic Algorithm(SGA) proposed by J. H. Holland is a population-based optimization method based on the principle of the Darwinian natural selection. The theoretical foundations of GA are the Schema Theorem and the Building Block Hypothesis. Although GA does well in many applications as an optimization method, still it does not guarantee the convergence to a global optimum in some problems. In designing intelligent systems, specially, since there is no deterministic solution, a heuristic trial-and error procedure is usually used to determine the systems' parameters. As an alternative scheme, therefore, there is a growing interest in a co-evolutionary system, where two populations constantly interact and co-evolve. In this paper we review the existing co-evolutionary algorithms and propose co-evolutionary schemes designing intelligent systems according to the relation between the system's components.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.26 no.4
• /
• pp.63-71
• /
• 2003

This paper analyses a deterministic scheduling problem concerned with manufacturing multiple types of components at a pre-assembly stage composed of parallel fabrication machines. Each component part is machined on a fabrication machine specified in advance. The manufactured components are subsequently assembled into products. The completion time of a job(product) is measured by the latest completion time of its all components at the pre-assembly stage. The problem has the objective measure of minimizing the total weighted completion time of a finite number of jobs. Two lower bounds are derived and tested in a branch-and-bound scheme. Also, three constructive heuristic algorithms are developed based on the machine aggregation and greedy strategies. Some empirical evaluation of the performance of the proposed branch-and-bound and heuristic algorithms are also performed.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.10
• /
• pp.23-31
• /
• 2018

Recently, there have been many studies on machine learning. Among them, studies on reinforcement learning are actively worked. In this study, we propose a controller to control bicycle using DDPG (Deep Deterministic Policy Gradient) algorithm which is the latest deep reinforcement learning method. In this paper, we redefine the compensation function of bicycle dynamics and neural network to learn agents. When using the proposed method for data learning and control, it is possible to perform the function of not allowing the bicycle to fall over and reach the further given destination unlike the existing method. For the performance evaluation, we have experimented that the proposed algorithm works in various environments such as fixed speed, random, target point, and not determined. Finally, as a result, it is confirmed that the proposed algorithm shows better performance than the conventional neural network algorithms NAF and PPO.