• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.7
• /
• pp.179-189
• /
• 1994

A two-phase neural network finds exact feasible solutions for a constrained optimization programming problem. The time-varying programming neural network is a modified steepest-gradient algorithm which solves time-varying optimization problems. In this paper, we propose a time-varying two-phase optimization neural network which incorporates the merits of the two-phase neural network and the time-varying neural network. The proposed algorithm is applied to system identification and function approximation using a multi-layer perceptron. Particularly training of a multi-layer perceptrion is regarded as a time-varying optimization problem. Our algorithm can also be applied to the case where the weights are constrained. Simulation results prove the proposed algorithm is efficient for solving various optimization problems.

• 유체기계공업학회:학술대회논문집
• /
• 2005.12a
• /
• pp.240-243
• /
• 2005

This study deals with the shape optimization of a wire spacer fuel assembly of Liquid Metal Reactors (LMRs). The Response Surface based optimization Method is used as an optimization technique with the Reynolds-averaged Navier-Stokes analysis of fluid flow and heat transfer using Shear Stress Transport (SST) turbulence model as a turbulence closure. Two design variables namely, pitch to fuel rod diameter ratio and lead length to fuel rod diameter ratio are selected. The objective function is defined as a combination of the heat transfer rate and the inverse of friction loss with a weighting factor. Three level full-factorial method is used to determine the training points. In total, nine experiments have been performed numerically and the resulting datas have been analysed for optimization study. Also, a comparison has been made between the optimized surface and the reference one in this study.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.10
• /
• pp.240-247
• /
• 2020

The Korean military is having difficulty maintaining combat power due to the insufficient troop numbers caused by a demographic cliff and the reduction of the mandatory military service period. Recently, discussions on improving the military training system have increased significantly. This paper proposes an optimization model to establish military training schedules to improve combat power. The oblivion and learning effects on training tasks were quantified through a survey and applied to the model. The objective value, combat power, was calculated based on the total task scores of the unit members and the number of task failures after four weeks. The scenarios were configured by the change in educational conditions and initial scores of some tasks. As a result of scenario experiments, combat power has increased by at least 10% and up to 77%, which is sufficient to maintain combat power considering the change in troops. In addition, the planning of combat skill tasks has a significant impact on combat power. Through this research model, it is expected that military training managers will be able to establish a training schedule that maintains or improves the combat power of troops effectively.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2000.04b
• /
• pp.449-457
• /
• 2000

The paper describes the study of global approximate optimization utilizing soft computing techniques such as genetic algorithms (GA's), neural networks (NN's), and fuzzy inference systems(FIS). GA's provide the increasing probability of locating a global optimum over the entire design space associated with multimodality and nonlinearity. NN's can be used as a tool for function approximations, a rapid reanalysis model for subsequent use in design optimization. FIS facilitates to handle the quantitative design information under the case where the training data samples are not sufficiently provided or uncertain information is included in design modeling. Properties of soft computing techniques affect the quality of global approximate model. Evolutionary fuzzy modeling (EFM) and adaptive neuro-fuzzy inference system (ANFIS) are briefly introduced for structural optimization problem in this context. The paper presents the success of EFM depends on how optimally the fuzzy membership parameters are selected and how fuzzy rules are generated.

• The KSFM Journal of Fluid Machinery
• /
• v.12 no.4
• /
• pp.44-53
• /
• 2009

Numerical design optimization of a fan-shaped hole for film-cooling has been carried out to improve film-cooling effectiveness by combining a three-dimensional Reynolds-averaged Navier-Stokes analysis with the radial basis neural network method, a well known surrogate modeling technique for optimization. The injection angle of hole, lateral expansion angle of hole and ratio of length-to-diameter of the hole are chosen as design variables and spatially averaged film-cooling effectiveness is considered as an objective function which is to be maximized. Twenty training points are obtained by Latin Hypercube sampling for three design variables. Sequential quadratic programming is used to search for the optimal point from the constructed surrogate. The film-cooling effectiveness has been successfully improved by the optimization with increased value of all design variables as compared to the reference geometry.

• Wind and Structures
• /
• v.34 no.3
• /
• pp.259-274
• /
• 2022

The present paper is focused on analyzing a set of Computational Fluid Dynamics (CFD) simulation data on reducing orthogonal peak base moment coefficients on a high-rise rectangular building with wings. The study adopts an aerodynamic optimization procedure (AOP) composed of CFD, artificial neural network (ANN), and genetic algorithm (G.A.). A parametric study is primarily accomplished by altering the wing positions with 3D transient CFD analysis using k - ε turbulence models. The CFD technique is validated by taking up a wind tunnel test. The required design parameters are obtained at each design point and used for training ANN. The trained ANN models are used as surrogates to conduct optimization studies using G.A. Two single-objective optimizations are performed to minimize the peak base moment coefficients in the individual directions. An additional multiobjective optimization is implemented with the motivation of diminishing the two orthogonal peak base moments concurrently. Pareto-optimal solutions specifying the preferred building shapes are offered.

• Journal of Information Processing Systems
• /
• v.14 no.5
• /
• pp.1075-1086
• /
• 2018

In this paper a cross-validation algorithm for training probabilistic neural networks (PNNs) is presented in order to be applied to automatic face identification. Actually, standard PNNs perform pretty well for small and medium sized databases but they suffer from serious problems when it comes to using them with large databases like those encountered in biometrics applications. To address this issue, we proposed in this work a new training algorithm for PNNs to reduce the hidden layer's size and avoid over-fitting at the same time. The proposed training algorithm generates networks with a smaller hidden layer which contains only representative examples in the training data set. Moreover, adding new classes or samples after training does not require retraining, which is one of the main characteristics of this solution. Results presented in this work show a great improvement both in the processing speed and generalization of the proposed classifier. This improvement is mainly caused by reducing significantly the size of the hidden layer.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.4
• /
• pp.201-208
• /
• 1996

This paper proposes an efficient method for improving the training performance of the neural network using a hybrid of backpropagation algorithm and dynamic tunneling system.The backpropagation algorithm, which is the fast gradient descent method, is applied for high-speed optimization. The dynamic tunneling system, which is the deterministic method iwth a tunneling phenomenone, is applied for blobal optimization. Converging to the local minima by using the backpropagation algorithm, the approximate initial point for escaping the local minima is estimated by the pattern classification, and the simulation results show that the performance of proposed method is superior th that of backpropagation algorithm with randomized initial point settings.

• Journal of Information Processing Systems
• /
• v.13 no.5
• /
• pp.1345-1357
• /
• 2017

A prototype selection method chooses a small set of training points from a whole set of class data. As the data size increases, the selected prototypes play a significant role in covering class regions and learning a discriminate rule. This paper discusses the methods for selecting prototypes in a classification framework. We formulate a prototype selection problem into a set covering optimization problem in which the sets are composed with distance metric and predefined classes. The formulation of our problem makes us draw attention only to prototypes per class, not considering the other class points. A training point becomes a prototype by checking the number of neighbors and whether it is preselected. In this setting, we propose a greedy algorithm which chooses the most relevant points for preserving the class dominant regions. The proposed method is simple to implement, does not have parameters to adapt, and achieves better or comparable results on both artificial and real-world problems.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.2226-2229
• /
• 2003

A new method is presented to construct a classifier. This was accomplished by combining a generalized regression neural network (GRNN) and a genetic algorithm (GA). The classifier constructed in this way is referred to as a GA-GRNN. The GA played a role of controlling training factors simultaneously. In GA optimization, neuron spreads were represented in a chromosome. The proposed optimization method was applied to a data set, consisted of 4 different promoter sequences. The training and test data were composed of 115 and 58 sequence patterns, respectively. The range of neuron spreads was experimentally varied from 0.4 to 1.4 with an increment of 0.1. The GA-GRNN was compared to a conventional GRNN. The classifier performance was investigated in terms of the classification sensitivity and prediction accuracy. The GA-GRNN significantly improved the total classification sensitivity compared to the conventional GRNN. Also, the GA-GRNN demonstrated an improvement of about 10.1% in the total prediction accuracy. As a result, the proposed GA-GRNN illustrated improved classification sensitivity and prediction accuracy over the conventional GRNN.