• STI Policy Review
• /
• v.2 no.2
• /
• pp.45-54
• /
• 2011

This paper develops a recursive multinomial probit model and describes its estimation method. The recursive multinomial probit model is an extension of a recursive bivariate probit model. The main difference between the two models is that a single decision among two or more alternatives can be considered in each choice equation in the proposed model. The recursive multinomial probit model is developed based on a standard framework of the multinomial probit model and a Bayesian approach with a Gibbs sampling is adopted for the estimation. The simulation exercise with artificial data sets is showed that the model performed well. Since the recursive multinomial probit model can be applied to analyze the causal relationship between discrete dependent variables with more than two outcomes, the model can play an important role in extending the methodology of the causal relationship analysis in innovation research.

• The Journal of Information Technology
• /
• v.9 no.3
• /
• pp.77-89
• /
• 2006

The event study analyzes returns around event date at a time. Event study provides estimation periods and cumulative returns. Stock split announcements are generally associated with positive abnormal returns. It is important to investigate the responses of stocks to new information contained in the announcements of stock splits. So It is important to study the long term performance in the case of Stock Split. This Study forced to two approach method in evaluating the performance, the event time portfolio approach and calendar time portfolio approach. The event time portfolio approach exists the CAR model, BHAR model and WR model. And the calendar time portfolio approach has the 3 factor model, 4 factor model, CTAR model, and RATS model.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.4
• /
• pp.2093-2108
• /
• 2017

In this paper, we describe a novel method for recognizing human actions from different views via view knowledge transfer. Our approach is characterized by two aspects: 1) We propose a unsupervised topic transfer model (TTM) to model two view-dependent vocabularies, where the original bag of visual words (BoVW) representation can be transferred into a bag of topics (BoT) representation. The higher-level BoT features, which can be shared across views, can connect action models for different views. 2) Our features make it possible to obtain a discriminative model of action under one view and categorize actions in another view. We tested our approach on the IXMAS data set, and the results are promising, given such a simple approach. In addition, we also demonstrate a supervised topic transfer model (STTM), which can combine transfer feature learning and discriminative classifier learning into one framework.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.17 no.2
• /
• pp.77-83
• /
• 2022

In this paper, the two-stage object detection approach is proposed to implement a deep learning-based label inspection system on edge computing environments. Since the label printed on the products during the production process contains important information related to the product, it is significantly to check the label information is correct. The proposed system uses the lightweight deep learning model that able to employ in the low-performance edge computing devices, and the two-stage object detection approach is applied to compensate for the low accuracy relatively. The proposed Two-Stage object detection approach consists of two object detection networks, Label Area Detection Network and Character Detection Network. Label Area Detection Network finds the label area in the product image, and Character Detection Network detects the words in the label area. Using this approach, we can detect characters precise even with a lightweight deep learning models. The SF-YOLO model applied in the proposed system is the YOLO-based lightweight object detection network designed for edge computing devices. This model showed up to 2 times faster processing time and a considerable improvement in accuracy, compared to other YOLO-based lightweight models such as YOLOv3-tiny and YOLOv4-tiny. Also since the amount of computation is low, it can be easily applied in edge computing environments.

• ETRI Journal
• /
• v.37 no.1
• /
• pp.175-185
• /
• 2015

This paper presents a system-of-systems (SoS) approach to the formal modeling of a cyber-physical system (CPS) for simulation-based analysis. The approach is based on a convergence technology for modeling and simulation of a highly complex system in which SoS modeling methodology, hybrid systems modeling theory, and simulation interoperation technology are merged. The methodology maps each constituent system of a CPS to a disparate model of either continuous or discrete types. The theory employs two formalisms for modeling of the two model types with formal specification of interfaces between them. Finally, the technology adapts a simulation bus called DEVS BUS whose protocol synchronizes time and exchange messages between subsystems simulation. Benefits of the approach include reusability of simulation models and environments, and simulation-based analysis of subsystems of a CPS in an inter-relational manner.

• Communications for Statistical Applications and Methods
• /
• v.30 no.6
• /
• pp.631-652
• /
• 2023

In the two-condition within-subject mediation design, pairs of variables such as mediator and outcome are observed under two treatment conditions. The main objective of the design is to investigate the indirect effects of the condition difference (sum) on the outcome difference (sum) through the mediator difference (sum) for comparison of two treatment conditions. The natural condition variables mean the original variables, while the rotated condition variables mean the difference and the sum of two natural variables. The outcome difference (sum) is expressed as a linear model regressed on two natural (rotated) mediators as a parallel two-mediator design in two condition approaches: the natural condition approach uses regressors as the natural condition variables, while the rotated condition approach uses regressors as the rotated condition variables. In each condition approach, the total indirect effect on the outcome difference (sum) can be expressed as the sum of two individual indirect effects: within- and cross-condition indirect effects. The total indirect effects on the outcome difference (sum) for both condition approaches are the same. The invariance of the total indirect effect makes it possible to analyze the nature of two pairs of individual indirect effects induced from the natural conditions and the rotated conditions. The two-condition within-subject design is extended to the addition of a between-subject moderator. Probing of the conditional indirect effects given the moderator values is implemented by plotting the bootstrap confidence intervals of indirect effects against the moderator values. The expected indirect effect with respect to the moderator is derived to provide the overall effect of moderator on the indirect effect. The model coefficients are estimated by the structural equation modeling approach and their statistical significance is tested using the bias-corrected bootstrap confidence intervals. All procedures are evaluated using function lavaan() of package {lavaan} in R.

• Food Science of Animal Resources
• /
• v.43 no.2
• /
• pp.374-381
• /
• 2023

In a previous study, 'response surface methodology (RSM) using a fullest balanced model' was proposed to improve the optimization of food processing when a standard second-order model has a significant lack of fit. However, that methodology can be used when each factor of the experimental design has five levels. In response surface experiments for optimization, not only five-level designs, but also three-level designs are used. Therefore, the present study aimed to improve the optimization of food processing when the experimental factors have three levels through a new approach to RSM. This approach employs three-step modeling based on a second-order model, a balanced higher-order model, and a balanced highest-order model. The dataset from the experimental data in a three-level, two-factor central composite design in a previous research was used to illustrate three-step modeling and the subsequent optimization. The proposed approach to RSM predicted improved results of optimization, which are different from the predicted optimization results in the previous research.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.10
• /
• pp.136-152
• /
• 2017

The thermodynamic models, PC-SAFT (Perturbed-Chain Statistical Associated Fluid Theory) state equation and the Two-model approach liquid activity coefficient model NRTL (Non Random Two Liquid) + Henry + Peng-Robinson, for modeling the Rectisol process using methanol aqueous solution as the $CO_2$ removal solvent were compared. In addition, to determine the new binary interaction parameters of the PC-SAFT state equations and the Henry's constant of the two-model approach, absorption equilibrium experiments between carbon dioxide and methanol at 273.25K and 262.35K were carried out and regression analysis was performed. The accuracy of the newly determined parameters was verified through the regression results of the experimental data. These model equations and validated parameters were used to model the carbon dioxide removal process. In the case of using the two-model approach, the methanol solvent flow rate required to remove 99.00% of $CO_2$ was estimated to be approximately 43.72% higher, the cooling water consumption in the distillation tower was 39.22% higher, and the steam consumption was 43.09% higher than that using PC-SAFT EOS. In conclusion, the Rectisol process operating under high pressure was designed to be larger than that using the PC-SAFT state equation when modeled using the liquid activity coefficient model equation with Henry's relation. For this reason, if the quantity of low-solubility gas components dissolved in a liquid at a constant temperature is proportional to the partial pressure of the gas phase, the carbon dioxide with high solubility in methanol does not predict the absorption characteristics between methanol and carbon dioxide.

• Smart Structures and Systems
• /
• v.29 no.5
• /
• pp.729-746
• /
• 2022

In this paper, a nonlinear structural model updating methodology based on the Deep Belief Network (DBN) is proposed. Firstly, the instantaneous parameters of the vibration responses are obtained by the discrete analytical mode decomposition (DAMD) method and the Hilbert transform (HT). The instantaneous parameters are regarded as the independent variables, and the nonlinear model parameters are considered as the dependent variables. Then the DBN is utilized for approximating the nonlinear mapping relationship between them. At last, the instantaneous parameters of the measured vibration responses are fed into the well-trained DBN. Owing to the strong learning and generalization abilities of the DBN, the updated nonlinear model parameters can be directly estimated. Two nonlinear shear-type structure models under two types of excitation and various noise levels are adopted as numerical simulations to validate the effectiveness of the proposed approach. The nonlinear properties of the structure model are simulated via the hysteretic parameters of a Bouc-Wen model and a Giuffré-Menegotto-Pinto model, respectively. Besides, the proposed approach is verified by a three-story shear-type frame with a piezoelectric friction damper (PFD). Simulated and experimental results suggest that the nonlinear model updating approach has high computational efficiency and precision.

• 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%.