• Journal of Information Science Theory and Practice
• /
• v.2 no.2
• /
• pp.16-32
• /
• 2014

As Topic Modeling has been applied in increasingly various domains, the difficulty in naming and characterizing topics also has been recognized more. This study, therefore, explores an approach of combining text mining with network analysis in a multi-step approach. The concept of synchronization was applied to re-assign the top author keywords in more than one topic category, in order to improve the visibility of the topic-author keyword network, and to increase the topical cohesion in each topic. The suggested approach was applied using 16,548 articles with 2,881 unique author keywords in construction and building engineering indexed by KSCI. As a result, it was revealed that the combined approach could improve both the visibility of the topic-author keyword map and topical cohesion in most of the detected topic categories. There should be more cases of applying the approach in various domains for generalization and advancement of the approach. Also, more sophisticated evaluation methods should also be necessary to develop the suggested approach.

• Proceedings of the KIEE Conference
• /
• 1998.11b
• /
• pp.537-539
• /
• 1998

In this paper, we construct the on-line model structure for the nonlinear process systems using the adaptive fuzzy-neural network. Adaptive fuzzy-neural network usually consists of two distinct modifiable structure, with both, the premise and the consequent part. These two parts can be adapted by different optimization methods, which are the hybrid learning procedure combining gradient descent method and least square method. To achieve the on-line model structure, we use the recursive least square method for the consequent parameter identification of nonlinear process. We design the interface between PLC and main computer, and construct the monitoring and control simulator for the nonlinear process. The proposed on-line modeling to real process is carried out to obtain the effective and accurate results.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2006.06a
• /
• pp.426-427
• /
• 2006

In this paper, the neural network model of the pretilt angle in the nematic liquid crystal on the homogeneous polyimide surface with different surface treatments is investigated. The pretilt angle is one of the main factors to determine the alignment of the liquid crystal display. The pretilt angle is measured to analyze the variation of the characteristics on the various process conditions. The rubbing strength and the hard baking temperature are considered as input factors. Latin hypercube sampling was used to generate initial weights and biases.

• KOREAN JOURNAL OF PACKAGING SCIENCE & TECHNOLOGY
• /
• v.29 no.3
• /
• pp.163-174
• /
• 2023

Though box compression strength (BCS) is commonly used as a performance criterion for shipping containers, estimating BCS remains a challenge. In this study, artificial neural networks (ANN) are implemented as a new tool, with a focus on building up ANN architectures for BCS estimation. An Artificial Neural Network (ANN) model can be constructed by adjusting four modeling factors: hidden neuron numbers, epochs, number of modeling cycles, and number of data points. The four factors interact with each other to influence model accuracy and can be optimized by minimizing model's Mean Squared Error (MSE). Using both data from the literature and "synthetic" data based on the McKee equation, we find that model estimation accuracy remains limited due to the uncertainty in both the input parameters and the ANN process itself. The population size to build an ANN model has been identified based on different data sets. This study provides a methodology guide for future research exploring the applicability of ANN to address problems and answer questions in the corrugated industry.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.34 no.6
• /
• pp.42-50
• /
• 2006

This paper addresses the optimized thermal modeling methodology for spacecraft board level thermal analysis. A direct thermal modeling of external and internal structure of active parts which have high power dissipation is newly proposed, based on conventional plate modeling for Printed Circuit Board(PCB). The parts thermal modeling results were compared with other generic methodologies and verified by thermal vacuum test. This parts thermal modeling was directly applied to thermal analysis of CS(Communication Subsystem) board of HAUSAT-2 small satellite. As a result, it was confirmed that the parts thermal modeling can complement other conventional modeling methodologies. A parts thermal modeling is very effective for thermal control design, since the existing thermal problems can be solved at the parts level in advance.

• The KIPS Transactions:PartC
• /
• v.11C no.2
• /
• pp.269-276
• /
• 2004

There exist several types of distributed subscriber networks using Asymmetric Digital Subscriber Line(ADSL), Very high -bit rate Digital subscriber Line(VDSL), and Data Oner Cable Service Interface Specifications(DOCSIS). The efficient and concentrated network management of those several distributed subscribers networks and resources requires the general management information model of network, which has abstract and conceptual managed objects of the heterogeneous networks and its equipment to manage the integrated subscriber network. This paper presents the general Internet subscribers network modeling framework using RM-ODP to manage that network in the form of integrated hierarchy. This paper adopts the object-oriented development methodology with UML and designs and implements the HFC network of DOCSIS as an example of the subscriber network.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.17 no.5
• /
• pp.73-86
• /
• 2007

Numerous types of models have been developed in recent years in response to the cyber threat posed by worms in order to analyze their propagation and predict their spread. Some of the most important ones involve mathematical modeling techniques such as Epidemic, AAWP (Analytical Active Worm Propagation Modeling) and LAAWP (Local AAWP). However, most models have several inherent limitations. For instance, they target worms that employ random scanning in the entire nv4 network and fail to consider the effects of countermeasures, making it difficult to analyze the extent of damage done by them and the effects of countermeasures in a specific network. This paper extends the equations and parameters of AAWP and LAAWP and suggests ALAAWP (Advanced LAAWP), a new worm simulation technique that rectifies the drawbacks of existing models.

• International Journal of Computer Science & Network Security
• /
• v.21 no.5
• /
• pp.139-150
• /
• 2021

Object orientation has become the predominant paradigm for conceptual modeling (e.g., UML), where the notions of class and object form the primitive building blocks of thought. Classes act as templates for objects that have attributes and methods (actions). The modeled systems are not even necessarily software systems: They can be human and artificial systems of many different kinds (e.g., teaching and learning systems). The UML class diagram is described as a central component of model-driven software development. It is the most common diagram in object-oriented models and used to model the static design view of a system. Objects both carry data and execute actions. According to some authorities in modeling, a certain degree of difficulty exists in understanding the semantics of these notions in UML class diagrams. Some researchers claim class diagrams have limited use for conceptual analysis and that they are best used for logical design. Performing conceptual analysis should not concern the ways facts are grouped into structures. Whether a fact will end up in the design as an attribute is not a conceptual issue. UML leads to drilling down into physical design details (e.g., private/public attributes, encapsulated operations, and navigating direction of an association). This paper is a venture to further the understanding of object-orientated concepts as exemplified in UML with the aim of developing a broad comprehension of conceptual modeling fundamentals. Thinging machine (TM) modeling is a new modeling language employed in such an undertaking. TM modeling interlaces structure (components) and actionality where actions infiltrate the attributes as much as the classes. Although space limitations affect some aspects of the class diagram, the concluding assessment of this study reveals the class description is a kind of shorthand for a richer sematic TM construct.

• Geophysics and Geophysical Exploration
• /
• v.26 no.3
• /
• pp.114-125
• /
• 2023

The physics-informed neural network (PINN) has been proposed to overcome the limitations of various numerical methods used to solve partial differential equations (PDEs) and the drawbacks of purely data-driven machine learning. The PINN directly applies PDEs to the construction of the loss function, introducing physical constraints to machine learning training. This technique can also be applied to wave equation modeling. However, to solve the wave equation using the PINN, second-order differentiations with respect to input data must be performed during neural network training, and the resulting wavefields contain complex dynamical phenomena, requiring careful strategies. This tutorial elucidates the fundamental concepts of the PINN and discusses considerations for wave equation modeling using the PINN approach. These considerations include spatial coordinate normalization, the selection of activation functions, and strategies for incorporating physics loss. Our experimental results demonstrated that normalizing the spatial coordinates of the training data leads to a more accurate reflection of initial conditions in neural network training for wave equation modeling. Furthermore, the characteristics of various functions were compared to select an appropriate activation function for wavefield prediction using neural networks. These comparisons focused on their differentiation with respect to input data and their convergence properties. Finally, the results of two scenarios for incorporating physics loss into the loss function during neural network training were compared. Through numerical experiments, a curriculum-based learning strategy, applying physics loss after the initial training steps, was more effective than utilizing physics loss from the early training steps. In addition, the effectiveness of the PINN technique was confirmed by comparing these results with those of training without any use of physics loss.

• Structural Engineering and Mechanics
• /
• v.61 no.2
• /
• pp.283-293
• /
• 2017

The main contribution of the present paper is to propose an intelligent fuzzy inference system approach for modeling the debonding strength of masonry elements retrofitted with Fiber Reinforced Polymer (FRP). To achieve this, the hybrid of meta-heuristic optimization methods and adaptive-network-based fuzzy inference system (ANFIS) is implemented. In this study, particle swarm optimization with passive congregation (PSOPC) and real coded genetic algorithm (RCGA) are used to determine the best parameters of ANFIS from which better bond strength models in terms of modeling accuracy can be generated. To evaluate the accuracy of the proposed PSOPC-ANFIS and RCGA-ANFIS approaches, the numerical results are compared based on a database from laboratory testing results of 109 sub-assemblages. The statistical evaluation results demonstrate that PSOPC-ANFIS in comparison with ANFIS-RCGA considerably enhances the accuracy of the ANFIS approach. Furthermore, the comparison between the proposed approaches and other soft computing methods indicate that the approaches can effectively predict the debonding strength and that their modeling results outperform those based on the other methods.