• 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 the Korean Statistical Society
• /
• v.29 no.1
• /
• pp.9-16
• /
• 2000

This article addresses the problem of maximum likelihood estimation in ARCH regression with lagged dependent variables. Some topics in asymptotics of the model such as uniform expansion of likelihood function and construction of a class of MLE are discussed, and the regularity property of MLE is obtained. The error process here is possibly non-Gaussian.

• The KIPS Transactions:PartB
• /
• v.15B no.6
• /
• pp.561-574
• /
• 2008

The spot detection phase of the 2-DE image analysis program segments a gel image into spot regions by an image segmentation algorithm and fits the spot regions to a spot shape model and quantifies the spot informations for the next phases. Currently the watershed algorithm is generally used as the segmentation algorithm and there are the Gaussian model and the diffusion model for the shape model. The diffusion model is closer to real spot shapes than the Gaussian model however spots have very various shapes and especially an asymmetric formation in x-coordinate and y-coordinate. The reason for asymmetric formation of spots is known that a protein could not be diffused completely because the 2-DE could not be processed under the ideal environment usually. Accordingly we propose an asymmetric diffusion model in this paper. The asymmetric diffusion model assumes that a protein spot is diffused from a disc at initial time of diffusing process, but is diffused asymmetrically for x-axis and y-axis respectively as time goes on. In experiments we processed spot matching for 19 gel images by using three models respectively and evaluated averages of SNR for comparing three models. As averages of SNR we got 14.22dB for the Gaussian model, 20.72dB for the diffusion model and 22.85dB for the asymmetric diffusion model. By experimental results we could confirm the asymmetric diffusion model is more efficient and more adequate for spot matching than the Gaussian model and the diffusion model.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.11
• /
• pp.1098-1102
• /
• 2014

Localization of aerial vehicles and map building of flight environments are key technologies for the autonomous flight of small UAVs. In outdoor environments, an unmanned aircraft can easily use a GPS (Global Positioning System) for its localization with acceptable accuracy. However, as the GPS is not available for use in indoor environments, the development of a SLAM (Simultaneous Localization and Mapping) system that is suitable for small UAVs is therefore needed. In this paper, we suggest a vision-based SLAM system that uses vision sensors and an AHRS (Attitude Heading Reference System) sensor. Feature points in images captured from the vision sensor are obtained by using GPU (Graphics Process Unit) based SIFT (Scale-invariant Feature Transform) algorithm. Those feature points are then combined with attitude information obtained from the AHRS to estimate the position of the small UAV. Based on the location information and color distribution, a Gaussian process model is generated, which could be a map. The experimental results show that the position of a small unmanned aircraft is estimated properly and the map of the environment is constructed by using the proposed method. Finally, the reliability of the proposed method is verified by comparing the difference between the estimated values and the actual values.

• Geomechanics and Engineering
• /
• v.28 no.4
• /
• pp.359-374
• /
• 2022

Geology conditions are crucial in decision-making during the planning and design phase of a tunnel project. Estimation of the geology conditions of road tunnels is subject to significant uncertainties. In this work, the effectiveness of a novel regression method in estimating geological or geotechnical parameters of road tunnel projects was explored. This method, called Gaussian process regression (GPR), formulates the learning of the regressor within a Bayesian framework. The GPR model was trained with data of old tunnel projects. To verify its feasibility, the GPR technique was applied to a road tunnel to predict the state of three geological/geomechanical parameters of Rock Mass Rating (RMR), Rock Structure Rating (RSR) and Q-value. Finally, in order to validate the GPR approach, the forecasted results were compared to the field-observed results. From this comparison, it was concluded that, the GPR is presented very good predictions. The R-squared values between the predicted results of the GPR vs. field-observed results for the RMR, RSR and Q-value were obtained equal to 0.8581, 0.8148 and 0.8788, respectively.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.28 no.3
• /
• pp.617-623
• /
• 2018

This study focuses on detection of malicious code through AndroidManifest permissoion feature extracted based on Android static analysis. Features are built on the permissions of AndroidManifest, which can save resources and time for analysis. Malicious app detection model consisted of SVM (support vector machine), NB (Naive Bayes), Gradient Boosting Classifier (GBC) and Logistic Regression model which learned 1,500 normal apps and 500 malicious apps and 98% detection rate. In addition, malicious app family identification is implemented by multi-classifiers model using algorithm SVM, GPC (Gaussian Process Classifier) and GBC (Gradient Boosting Classifier). The learned family identification machine learning model identified 92% of malicious app families.

• Journal of Korea Multimedia Society
• /
• v.11 no.5
• /
• pp.597-609
• /
• 2008

Due to the recursive updating nature of background model, previous background modeling methods are often perturbed by crowd scenes where foreground pixels occurs more frequently than background pixels. To resolve this problem, an adaptive background modeling method, which is based on the well-known Gaussian mixture background model, is proposed. In the proposed method, the learning rate of background model is adaptively adjusted with respect to the crowdedness of the scene. Consequently, the learning process is suppressed in crowded scene to maintain proper background model. Experiments on real dataset revealed that the proposed method could perform background subtraction effectively even in crowd situation while the performance is almost the same to the previous method in normal scenes. Also, the F-measure was increased by 5-10% compared to the previous background modeling methods in the video of crowded situations.

• Journal of IKEEE
• /
• v.20 no.1
• /
• pp.94-102
• /
• 2016

This paper proposes a new prediction method to reduce times and labor of repetitive multi-physics simulation. To achieve exact results from the whole simulation processes, complex modeling and huge amounts of time are required. Current multi-physics analysis focuses on the simulation method itself and the simulation environment to reduce times and labor. However this paper proposes an alternative way to reduce simulation times and labor by exploiting machine learning algorithm trained with data set from simulation results. Through comparing each machine learning algorithm, Gaussian Process Regression showed the best performance with under 100 training data and how similar results can be achieved through machine-learning without a complex simulation process. Given trained machine learning algorithm, it's possible to predict the result after changing some features of the simulation model just in a few second. This new method will be helpful to effectively reduce simulation times and labor because it can predict the results before more simulation.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.3
• /
• pp.138-142
• /
• 2009

We propose an effective voice-phishing detection algorithm based on discriminative weight training. The detection of voice phishing is performed based on a Gaussian mixture model (GMM) incorporaiting minimum classification error (MCE) technique. Actually, the MCE technique is based on log-likelihood from the decoding parameter of the SMV(Selectable Mode Vocoder) directly extracted from the decoding process in the mobile phone. According to the experimental result, the proposed approach is found to be effective for the voice phishing detection.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.5
• /
• pp.415-419
• /
• 2008

Detecting and correcting defects on LCD glasses early in the manufacturing process becomes important for panel makers to reduce the manufacturing costs and to improve productivity. Many attempts have been made and were successfully applied to detect and identify simple defects such as scratches, dents, and foreign objects on glasses. However, it is still difficult to robustly detect low-contrast defect region, called Mura or blemish area on glasses. Typically, these defect areas are roughly defined as relatively large, several millimeters of diameter, and relatively dark and/or bright region of low Signal-to-Noise Ratio (SNR) against background of low-frequency signal. The aim of this article is to present a robust algorithm to segment these blemish defects. Early 90's, a highly robust estimator, known as the Model-Fitting (MF) estimator was developed by X. Zhuang et. al. and have been successfully used in many computer vision application. Compared to the conventional Least-Square (LS) estimator the MF estimator can successfully estimate model parameters from a dataset of contaminated Gaussian mixture. Such a noise model is defined as a regular white Gaussian noise model with probability $1-\varepsilon$ plus an outlier process with probability $varepsilon$. In the sense of robust estimation, the blemish defect in images can be considered as being a group of outliers in the process of estimating image background model parameters. The algorithm developed in this paper uses a modified MF estimator to robustly estimate the background model and as a by-product to segment the blemish defects, the outliers.