• Journal of Biomedical Engineering Research
• /
• v.28 no.6
• /
• pp.817-824
• /
• 2007

Tractography using Diffusion Tensor Magnetic Resonance Imaging (DT-MRI) is a method to determine the architecture of axonal fibers in the central nervous system by computing the direction of the principal eigenvector in the white matter of the brain. However, the fiber tracking methods suffer from the noise included in the diffusion tensor images that affects the determination of the principal eigenvector. As the fiber tracking progresses, the accumulated error creates a large deviation between the calculated fiber and the real fiber. This problem of the DT-MRI tractography is known mathematically as the ill-posed problem which means that tractography is very sensitive to perturbations by noise. To reduce the noise in DT-MRI measurements, a tensor-valued median filter which is reported to be denoising and structure-preserving in fiber tracking, is applied in the tractography. In this paper, we proposed the modified gradient descent method which converges fast and accurately to the optimal tensor-valued median filter by changing the step size. In addition, the performance of the modified gradient descent method is compared with others. We used the synthetic image which consists of 45 degree principal eigenvectors and the corticospinal tract. For the synthetic image, the proposed method achieved 4.66%, 16.66% and 15.08% less error than the conventional gradient descent method for error measures AE, AAE, AFA respectively. For the corticospinal tract, at iteration number ten the proposed method achieved 3.78%, 25.71 % and 11.54% less error than the conventional gradient descent method for error measures AE, AAE, AFA respectively.

• Journal of Digital Contents Society
• /
• v.18 no.2
• /
• pp.347-356
• /
• 2017

It is well-known that the face recognition method via sparse representation has been proved very robust and showed good performance. Its weakness is, however, that its time complexity is very high because it should solve $L_1$-minimization problem to find the sparse solution. In this paper, we propose to use the ROMP(Regularized Orthogonal Matching Pursuit) method for the sparse solution, which solves the $L_2$-minimization problem with regularization condition using the greed strategy. In experiments, we shows that the proposed method is comparable to the existing best $L_1$-minimization solver, Homotopy, but is 60 times faster than Homotopy. Also, we proposed C-SCI method for classification. The C-SCI method is very effective since it considers the sparse solution only without reconstructing the test data. It is shown that the C-SCI method is comparable to, but is 5 times faster than the existing best classification method.

• International Journal of Computer Science & Network Security
• /
• v.23 no.8
• /
• pp.177-189
• /
• 2023

Malware detection is an increasingly important operational focus in cyber security, particularly given the fast pace of such threats (e.g., new malware variants introduced every day). There has been great interest in exploring the use of machine learning techniques in automating and enhancing the effectiveness of malware detection and analysis. In this paper, we present a deep recurrent neural network solution as a stacked Long Short-Term Memory (LSTM) with a pre-training as a regularization method to avoid random network initialization. In our proposal, we use global and short dependencies of the inputs. With pre-training, we avoid random initialization and are able to improve the accuracy and robustness of malware threat hunting. The proposed method speeds up the convergence (in comparison to stacked LSTM) by reducing the length of malware OpCode or bytecode sequences. Hence, the complexity of our final method is reduced. This leads to better accuracy, higher Mattews Correlation Coefficients (MCC), and Area Under the Curve (AUC) in comparison to a standard LSTM with similar detection time. Our proposed method can be applied in real-time malware threat hunting, particularly for safety critical systems such as eHealth or Internet of Military of Things where poor convergence of the model could lead to catastrophic consequences. We evaluate the effectiveness of our proposed method on Windows, Ransomware, Internet of Things (IoT), and Android malware datasets using both static and dynamic analysis. For the IoT malware detection, we also present a comparative summary of the performance on an IoT-specific dataset of our proposed method and the standard stacked LSTM method. More specifically, of our proposed method achieves an accuracy of 99.1% in detecting IoT malware samples, with AUC of 0.985, and MCC of 0.95; thus, outperforming standard LSTM based methods in these key metrics.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.23 no.2
• /
• pp.176-184
• /
• 2013

Quantification of road noise is a challenging issue in the development of tire noise since its transfer paths are complicated. In this paper, a simplified model to estimate the road noise is developed. Transfer path of the model is from wheel to interior. The method uses the wheel excitation force estimated throughout inverse method. In inversion procedure, the Tikhonov regularization method is used to reduce the inversion error. To estimate the wheel excitation force, the vibration of knuckle is measured and transfer function between knuckle and wheel center is also measured. The wheel excitation force is estimated by using the measured knuckle vibration and the inversed transfer function. Finally interior noise due to wheel force is estimated by multiplying wheel excitation force in the vibro-acoustic transfer function. This vibro-acoustic transfer function is obtained throughout measurement. The proposed method is validated by using cleat excitation method. Finally, it is applied to the estimation of interior noise of the vehicle with different types of tires during driving test.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.4
• /
• pp.117-128
• /
• 2016

Installation of earth grounding system is essential to ensure personnel safety and correct operation of electrical equipment. Earth parameters, especially, soil resistivity has to be determined in designing an efficient earth grounding system. The most common applied technique to measure soil resistance is Wenner four-point method. Implementation of this method is expensive, time consuming and cumbersome as large set of measurements with variable electrode spacing are required to obtain a one dimensional resistivity plot. It is advantageous to have a method which is of low cost and provides fast measurements. In this perspective, electrical resistance tomography (ERT) is applied to estimate subsurface resistivity profile. Electrical resistance tomograms characterize the soil resistivity distribution based on the measurements from electrodes placed in the region of interest. The nonlinear ill-posed inverse problem is solved using iterated Gauss-Newton method with Tikhonov regularization. Through extensive numerical simulations, it is found that ERT offers promising performance in estimating the three-dimensional soil resistivity distribution.

• Computers and Concrete
• /
• v.20 no.5
• /
• pp.563-572
• /
• 2017

Ordinary reinforced concrete (RC) and prestressed concrete bridges are two popular and typical types of short- and medium-span bridges that accounts for the vast majority of all existing bridges. The cost of maintaining, repairing or replacing degraded existing RC bridges is immense. Detecting the abnormality of RC bridges at an early stage and taking the protective measures in advance are effective ways to improve maintenance practices and reduce the maintenance cost. This study proposes a systematic method from influence line (IL) identification to damage detection with applications to RC bridges. An IL identification method which integrates the cubic B-spline function with Tikhonov regularization is first proposed based on the vehicle information and the corresponding moving vehicle induced bridge response time history. Subsequently, IL change is defined as a damage index for bridge damage detection, and information fusion technique that synthesizes ILs of multiple locations/sensors is used to improve the efficiency and accuracy of damage localization. Finally, the feasibility of the proposed systematic method is verified through experimental tests on a three-span continuous RC beam. The comparison suggests that the identified ILs can well match with the baseline ILs, and it demonstrates that the proposed IL identification method has a high accuracy and a great potential in engineering applications. Results in this case indicate that deflection ILs are superior than strain ILs for damage detection of RC beams, and the performance of damage localization can be significantly improved with the information fusion of multiple ILs.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.5
• /
• pp.77-84
• /
• 2022

In this paper, we propose a novel unsupervised feature selection method. Conventional unsupervised feature selection method defines virtual label and uses a regression analysis that projects the given data to this label. However, since virtual labels are generated from data, they can be formed similarly in the space. Thus, in the conventional method, the features can be selected in only restricted space. To solve this problem, in this paper, features are selected using orthogonal projections and low-rank approximations. To solve this problem, in this paper, a virtual label is projected to orthogonal space and the given data set is also projected to this space. Through this process, effective features can be selected. In addition, projection matrix is restricted low-rank to allow more effective features to be selected in low-dimensional space. To achieve these objectives, a cost function is designed and an efficient optimization method is proposed. Experimental results for six data sets demonstrate that the proposed method outperforms existing conventional unsupervised feature selection methods in most cases.

• Journal of KIISE:Software and Applications
• /
• v.30 no.3_4
• /
• pp.326-338
• /
• 2003

The objective of a neural network design and model selection is to construct an optimal network with a good generalization performance. However, training data include noises, and the number of training data is not sufficient, which results in the difference between the true probability distribution and the empirical one. The difference makes the teaming parameters to over-fit only to training data and to deviate from the true distribution of data, which is called the overfitting phenomenon. The overfilled neural network shows good approximations for the training data, but gives bad predictions to untrained new data. As the complexity of the neural network increases, this overfitting phenomenon also becomes more severe. In this paper, by taking statistical viewpoint, we proposed an integrative process for neural network design and model selection method in order to improve generalization performance. At first, by using the natural gradient learning with adaptive regularization, we try to obtain optimal parameters that are not overfilled to training data with fast convergence. By adopting the natural pruning to the obtained optimal parameters, we generate several candidates of network model with different sizes. Finally, we select an optimal model among candidate models based on the Bayesian Information Criteria. Through the computer simulation on benchmark problems, we confirm the generalization and structure optimization performance of the proposed integrative process of teaming and model selection.

• Journal of the Korean Geotechnical Society
• /
• v.39 no.12
• /
• pp.37-46
• /
• 2023

Dynamic soil properties are essential factors for predicting the detailed behavior of the ground. However, there are limitations to gathering soil samples and performing additional experiments. In this study, we used an artificial neural network (ANN) to predict dynamic soil properties based on static soil properties. The selected static soil properties were soil cohesion, internal friction angle, porosity, specific gravity, and uniaxial compressive strength, whereas the compressional and shear wave velocities were determined for the dynamic soil properties. The Levenberg-Marquardt and Bayesian regularization methods were used to enhance the reliability of the ANN results, and the reliability associated with each optimization method was compared. The accuracy of the ANN model was represented by the coefficient of determination, which was greater than 0.9 in the training and testing phases, indicating that the proposed ANN model exhibits high reliability. Further, the reliability of the output values was verified with new input data, and the results showed high accuracy.

• Journal of Korea Multimedia Society
• /
• v.6 no.1
• /
• pp.136-147
• /
• 2003

To achieve an optimal performance for a given problem, we need an integrative process of the parameter optimization via learning and the structure optimization via model selection. In this paper, we propose an efficient optimization method for improving generalization performance by considering the property of each sub-method and by combining them with common theoretical properties. First, weight parameters are optimized by natural gradient teaming with adaptive regularization, which uses a diverse error function. Second, the network structure is optimized by eliminating unnecessary parameters with natural pruning. Through iterating these processes, candidate models are constructed and evaluated based on the Bayesian Information Criterion so that an optimal one is finally selected. Through computational experiments on benchmark problems, we confirm the weight parameter and structure optimization performance of the proposed method.