• ETRI Journal
• /
• v.33 no.6
• /
• pp.914-923
• /
• 2011

This paper concerns robust and reliable speaker model training for text-independent speaker verification. The baseline speaker modeling approach is the Gaussian mixture model (GMM). In text-independent speaker verification, the amount of speech data may be different for speakers. However, we still wish the modeling approach to perform equally well for all speakers. Besides, the modeling technique must be least vulnerable against unseen data. A traditional approach for GMM training is expectation maximization (EM) method, which is known for its overfitting problem and its weakness in handling insufficient training data. To tackle these problems, variational approximation is proposed. Variational approaches are known to be robust against overtraining and data insufficiency. We evaluated the proposed approach on two different databases, namely KING and TFarsdat. The experiments show that the proposed approach improves the performance on TFarsdat and KING databases by 0.56% and 4.81%, respectively. Also, the experiments show that the variationally optimized GMM is more robust against noise and the verification error rate in noisy environments for TFarsdat dataset decreases by 1.52%.

• Smart Structures and Systems
• /
• v.17 no.4
• /
• pp.611-629
• /
• 2016

This study presents a new approach of surrogate modeling for time-consuming finite element analysis. A surrogate model is widely used to reduce the computational cost under an iterative computational analysis. Although a variety of the methods have been widely investigated, there are still difficulties in surrogate modeling from a practical point of view: (1) How to derive optimal design of experiments (i.e., the number of training samples and their locations); and (2) diagnostics of the surrogate model. To overcome these difficulties, we propose a sequential surrogate modeling based on Gaussian process model (GPM) with self-adaptive sampling. The proposed approach not only enables further sampling to make GPM more accurate, but also evaluates the model adequacy within a sequential framework. The applicability of the proposed approach is first demonstrated by using mathematical test functions. Then, it is applied as a substitute of the iterative finite element analysis to Monte Carlo simulation for a response uncertainty analysis under correlated input uncertainties. In all numerical studies, it is successful to build GPM automatically with the minimal user intervention. The proposed approach can be customized for the various response surfaces and help a less experienced user save his/her efforts.

• Journal of Power Electronics
• /
• v.17 no.2
• /
• pp.411-421
• /
• 2017

An approach based on a 2D lumped model is presented to quantify the voltage transfer gain (VTG) in power converter low power planes. The advantage of the modeling approach is the ease with which typical noise reduction devices such as decoupling capacitors or ferrite beads can be integrated into the model. This feature is enforced by a new modular approach based on effective matrix partitioning, which is presented in the paper. This partitioning is used to decouple power plane equations from external device impedance, which avoids the need for rewriting of a whole set of equation at every change. The model is quickly solved in the frequency domain, which is well suited for an automated layout optimization algorithm. Using frequency domain modeling also allows the integration of frequency-dependent devices such inductors and capacitors, which are required for realistic computation results. In order to check the precision of the modeling approach, VTGs for several layout configurations are computed and compared with experimental measurements based on scattering parameters.

• ETRI Journal
• /
• v.39 no.5
• /
• pp.718-728
• /
• 2017

Empirical modeling of wireless fading channels using common schemes such as autoregression and the finite state Markov chain (FSMC) is investigated. The conceptual background of both channel structures and the establishment of their mutual dependence in a confined manner are presented. The novel contribution lies in the proposal of a new approach for deriving the state transition probabilities borrowed from economic disciplines, which has not been studied so far with respect to the modeling of FSMC wireless fading channels. The proposed approach is based on equal portioning of the received signal-to-noise ratio, realized by using an alternative probability construction that was initially highlighted by Tauchen. The associated statistical procedure shows that a first-order FSMC with a limited number of channel states can satisfactorily approximate fading. The computational overheads of the proposed technique are analyzed and proven to be less demanding compared to the conventional FSMC approach based on the level crossing rate. Simulations confirm the analytical results and promising performance of the new channel model based on the Tauchen approach without extra complexity costs.

• Korean Journal of Computational Design and Engineering
• /
• v.6 no.3
• /
• pp.157-168
• /
• 2001

The description of enterprise activities is the basis fur process improvement and information system building. To describe such activities, it is necessary to model the enterprise activities from the abstraction level to the implementation level in a stepwise and integrated form. For this reason, several modeling approaches have been proposed. However, most of them lacked the stepwise or integration aspects although some of them covered overall levels. This study proposes the hierarchical modeling approach for integrating the enterprise activity model from the abstraction level to the implementation level systematically. It is composed of five modeling levels such as function level, process level, task level, document workflow level, and event flow level. This study discusses the definition and characteristics of each level and compare our modeling frame with other modeling methodologies in case study.

• Journal of Ship and Ocean Technology
• /
• v.5 no.2
• /
• pp.1-13
• /
• 2001

Data in one Product Model for shipbuilding is inevitably referenced and manipulated during each phase of design or production activities, and data or manipulation status naturally varies from the original with the advance of each activities. For Object-Oriented approach, it is hard to identify classes dealing with those variations, and even if once a product model is developed, it might be getting much harder to modify it to cope with a new additional phase of activities. This paper proposes an Instance-Oriented Modeling Method, temporarily named “Concept-Relationship Modeling Approach”, which handles Data structure and Behavior independently of each other in order to resolve the difficulties above.

• 한국생물공학회:학술대회논문집
• /
• 2000.04a
• /
• pp.133-136
• /
• 2000

In recent years, hybrid neural network approaches which combine neural networks and mechanistic models have been gaining considerable interests. These approaches are potentially very efficient to obtain more accurate predictions of process dynamics by combining mechanistic and neural models in such a way that the neural network model properly captures unknown and nonlinear parts of the mechanistic model. In this work, such an approach was applied in the modeling of a full-scale coke wastewater treatment process. First, a simplified mechanistic model was developed based on the Activated Sludge Model No.1 and the specific process knowledge, Then neural network was incorporated with the mechanistic model to compensate the errors between the mechanistic model and the process data. Simulation and actual process data showed that the hybrid modeling approach could predict accurate process dynamics of industrial wastewater treatment plant. The promising results indicated that the hybrid modeling approach could be a useful tool for accurate and cost-effective modeling of biochemical processes.

• Journal of Environmental Science International
• /
• v.8 no.2
• /
• pp.211-219
• /
• 1999

Surface complexation models(SCMs) have been performed to predict metal ion adsorption behavior onto the mineral surface. Application of SCMs, however, requires a self-consistent approach to determine model parameter values. In this paper, in order to determine the metal ion adsorption parameters for the triple layer model(TLM) version of the SCM, we used the zeta potential data for Zeolite and Kaolinite, and the metal ion adsorption data for Pb(II) and Cd(II). Fitting parameters determined for the modeling were as follows ; total site concentration, site density, specific surface area, surface acidity constants, etc. Zeta potential as a new approach other than the acidic-alkalimetric titration method was adopted for simulation of adsorption phenomena. Some fitting parameters were determined by the trial and error method. Modeling approach was successful in quantitatively simulating adsorption behavior under various geochemical conditions.

• Korean Journal of Computational Design and Engineering
• /
• v.7 no.1
• /
• pp.42-56
• /
• 2002

Presented in this study is a top-down functional requirement analysis procedure and the desired functionalities for PDM system development, and the benefits of top-down approach over a conventional bottom-up approach is also shown. For the purpose of top-down requirement analysis for PDM system, this study proposes 4P modeling view. 4P modeling view is defined as a modeling perspective for classifying functional requirements and integrating product-related information objects that must be man-aged within PDM systems. Based on 4P modeling templates, benchmarking analysis of commercially major PDM products is conducted and as a result of this analysis, this study suggests desired functionalities for PDM system.

• International Journal of Computer Science & Network Security
• /
• v.23 no.7
• /
• pp.9-15
• /
• 2023

A common language for modeling software requirements and design in recent years is Unified Modeling Language (UML). Essential principles and rules are provided by UML to help visualize and comprehend complex software systems. It has therefore been incorporated into the curriculum for software engineering courses at several institutions all around the world. However, it is commonly recognized that UML is challenging for beginners to understand, mostly owing to its complexity and ill-defined nature. It is unavoidable that we need to comprehend their preferences and issues considerably better than we do presently to approach the problem of teaching UML to beginner students in an acceptable manner. This paper offers a hint-based approach that can be implemented along with an ordinary lab task. Some keywords are highlighted to indicate class diagram components and make students understand the textual descriptions. The experimental results indicate significant improvement in students' learning skills. Furthermore, the majority of students also positively responded to the survey conducted in the end experimental study.