• Journal of Digital Contents Society
• /
• v.18 no.5
• /
• pp.969-976
• /
• 2017

This paper proposes the learning method of an artificial neural network and a convolutional neural network using the WFSO algorithm developed as an optimization algorithm. Since the optimization algorithm searches based on a number of candidate solutions, it has a drawback in that it is generally slow, but it rarely falls into the local optimal solution and it is easy to parallelize. In addition, the artificial neural networks with non-differentiable activation functions can be trained and the structure and weights can be optimized at the same time. In this paper, we describe how to apply WFSO algorithm to artificial neural network learning and compare its performances with error back-propagation algorithm in multilayer artificial neural networks and convolutional neural networks.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.334-337
• /
• 2003

The emission of dioxins from waste incinerators is one of the most important environmental problems today, It is known that optimization of waste incinerator controllers is a very difficult problem due to the complex nature of the dynamic environment within the incinerator. In this paper, we propose applying artificial neural networks to waste incinerator controllers. We show that artificial neural networks can project the emission of dioxins with a fair degree of accuracy.

• Structural Engineering and Mechanics
• /
• v.52 no.4
• /
• pp.739-753
• /
• 2014

There have been many packages that can be employed to analyze plane frames. However, because most structural analysis packages suffer from closeness of system, it is very difficult to integrate it with an optimization package. To overcome the difficulty, we proposed a possible alternative, DAMDO, which integrate Design, Analysis, Modeling, Definition, and Optimization phases into an integrative environment. The DAMDO methodology employs neural networks to integrate structural analysis package and optimization package so as not to need directly to integrate these two packages. The key problem of the DAMDO approach is how to generate a set of reasonable random designs in the first phase. According to the characteristics of optimized plane frames, we proposed the ratio variable approach to generate them. The empirical results show that the ratio variable approach can greatly improve the accuracy of the neural networks, and the plane frame optimization problems can be solved by the DAMDO methodology.

• Geomechanics and Engineering
• /
• v.37 no.3
• /
• pp.263-277
• /
• 2024

Artificial Neural Networks (ANNs) are artificial learning algorithms that provide successful results in solving many machine learning problems such as classification, prediction, object detection, object segmentation, image and video classification. There is an increasing number of studies that use ANNs as a prediction tool in soil classification. The aim of this research was to understand the role of hyperparameter optimization in enhancing the accuracy of ANNs for soil type classification. The research results has shown that the hyperparameter optimization and hyperparamter optimized ANNs can be utilized as an efficient mechanism for increasing the estimation accuracy for this problem. It is observed that the developed hyperparameter tool (HyperNetExplorer) that is utilizing the Covariance Matrix Adaptation Evolution Strategy (CMAES), Genetic Algorithm (GA) and Jaya Algorithm (JA) optimization techniques can be successfully used for the discovery of hyperparameter optimized ANNs, which can accomplish soil classification with 100% accuracy.

• Computers and Concrete
• /
• v.14 no.4
• /
• pp.445-462
• /
• 2014

To solve structural optimization problems, it is necessary to integrate a structural analysis package and an optimization package. There have been many packages that can be employed to analyze reinforced concrete plane frames. However, because most structural analysis packages suffer from closeness of systems, it is very difficult to integrate them with optimization packages. To overcome the difficulty, we proposed a possible alternative, DAMDO, which integrates Design, Analysis, Modeling, Definition, and Optimization phases into an integration environment as follows. (1) Design: first generate many possible structural design alternatives. Each design alternative consists of many design variables X. (2) Analysis: employ the structural analysis software to analyze all structural design alternatives to obtain their internal forces and displacements. They are the response variables Y. (3) Modeling: employ artificial neural networks to build the models Y=f(X) to obtain the relationship functions between the design variables X and the response variables Y. (4) Definition: employ the design variables X and the response variables Y to define the objective function and constraint functions. (5) Optimization: employ the optimization software to solve the optimization problem consisting of the objective function and the constraint functions to produce the optimum design variables. The RC frame optimization problem was examined to evaluate the DAMDO approach, and the empirical results showed that it can be solved by the approach.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 1999.10a
• /
• pp.323-335
• /
• 1999

In this paper, we propose a new hybrid model of artificial neural networks(ANNs) and genetic algorithm (GA) to optimal feature transformation and feature weighting. Previous research proposed several variants of hybrid ANNs and GA models including feature weighting, feature subset selection and network structure optimization. Among the vast majority of these studies, however, ANNs did not learn the patterns of data well, because they employed GA for simple use. In this study, we incorporate GA in a simultaneous manner to improve the learning and generalization ability of ANNs. In this study, GA plays role to optimize feature weighting and feature transformation simultaneously. Globally optimized feature weighting overcome the well-known limitations of gradient descent algorithm and globally optimized feature transformation also reduce the dimensionality of the feature space and eliminate irrelevant factors in modeling ANNs. By this procedure, we can improve the performance and enhance the generalisability of ANNs.

• Communications for Statistical Applications and Methods
• /
• v.10 no.1
• /
• pp.153-166
• /
• 2003

This study presents the time-divided forecasting model to integrate evolutionary optimization algorithm and change point detection based on artificial neural networks (ANN) for the prediction of (Korea) stock price index. The genetic algorithm(GA) is introduced as an evolutionary optimization method in this study. The basic concept of the proposed model is to obtain intervals divided by change points, to identify them as optimal or near-optimal change point groups, and to use them in the forecasting of the stock price index. The proposed model consists of three phases. The first phase detects successive change points. The second phase detects the change-point groups with the GA. Finally, the third phase forecasts the output with ANN using the GA. This study examines the predictability of the proposed model for the prediction of stock price index.

• Advances in Computational Design
• /
• v.7 no.1
• /
• pp.1-17
• /
• 2022

In this study, a new strategy is presented to transmit the fundamental elastic beam problem into the modern optimization platform and solve it by using artificial intelligence (AI) tools. As a practical example, deflection of Euler-Bernoulli beam is mathematically formulated by 2nd-order ordinary differential equations (ODEs) in accordance to the classical beam theory. This fundamental engineer problem is then transmitted from classic formulation to its artificial-intelligence presentation where the behavior of the beam is simulated by using neural networks (NNs). The supervised training strategy is employed in the developed NNs implemented in the heuristic optimization algorithms as the fitness function. Different evolutionary optimization tools such as genetic algorithm (GA) and particle swarm optimization (PSO) are used to solve this non-linear optimization problem. The step-by-step procedure of the proposed method is presented in the form of a practical flowchart. The results indicate that the proposed method of using AI toolsin solving beam ODEs can efficiently lead to accurate solutions with low computational costs, and should prove useful to solve more complex practical applications.

• Computers and Concrete
• /
• v.6 no.3
• /
• pp.253-268
• /
• 2009

This study aims to develop a cost-based high-performance concrete (HPC) mix optimization system based on an integrated approach using artificial neural networks (ANNs) and genetic algorithms (GA). ANNs are used to predict the three main properties of HPC, namely workability, strength and durability, which are used to evaluate fitness and constraint violations in the GA process. Multilayer back-propagation neural networks are trained using the results obtained from experiments and previous research. The correlation between concrete components and its properties is established. GA is employed to arrive at an optimal mix proportion of HPC by minimizing its total cost. A system prototype, called High Performance Concrete Mix-Design System using Genetic Algorithm and Neural Networks (HPCGANN), was developed in MATLAB. The architecture of the proposed system consists of three main parts: 1) User interface; 2) ANNs prediction models software; and 3) GA engine software. The validation of the proposed system is carried out by comparing the results obtained from the system with the trial batches. The results indicate that the proposed system can be used to enable the design of HPC mix which corresponds to its required performance. Furthermore, the proposed system takes into account the influence of the fluctuating unit price of materials in order to achieve the lowest cost of concrete, which cannot be easily obtained by traditional methods or trial-and-error techniques.

• Proceedings of the KSME Conference
• /
• 2008.11a
• /
• pp.695-700
• /
• 2008

Artificial neural network (ANN) has been extensively used in areas of nonlinear system modeling, analysis and design applications. Basically, ANN has its distinct capabilities of implementing system identification and/or function approximation using a number of input/output patterns that can be obtained via numerical and/or experimental manners. The paper describes a role of ANN, especially a back-propagation neural network (BPN) in the context of engineering analysis, design and optimization. Fundamental mechanism of BPN is briefly summarized in terms of training procedure and function approximation. The BPN based causality analysis (CA) is further discussed to realize the problem decomposition in the context of multidisciplinary design optimization. Such CA is also applied to quantitatively evaluate the uncoupled or decoupled design matrix in the context of axiomatic design with the independence axiom.