• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.339-341
• /
• 1998

다층 퍼셉트론 학습은 학습 데이터의 능동적인 선택 여부에 따라 능동적 학습(Active learning)과 피동적 학습(Passive learning)으로 구분할 수 있다. 기존의 능동적 학습 방법들은 학습 데이터의 중요도를 측정할 수 있는 기준(measure)을 제시하고 이 기준에 따라 학습 데이터를 선택하는 방법을 취하고 있다. 이 방법들은 학습 데이터 선택을 위해 Hessian Approximation과 같은 복잡한 계산이나 학습 데이터를 선택하는 과정에 있어서 데이터의 중요도를 평가하기 위한 반복적인 계산을 필요로 한다. 본 논문에서는 학습 데이터 선택 시 반복적인 계산이 필요하지 않는 비교사 학습을 이용한 능동적 학습 데이터 선택 방법을 제안하고 그 수렴 특성과 일반화 성능을 분석한다. 또한 비교 실험을 통하여 제안된 방법이 기존의 능동적 학습방법보다 간단한 계산만으로 수렴 속도를 향상시키며 일반화에도 뒤떨어지지 않음을 보인다.

• IE interfaces
• /
• v.16 no.spc
• /
• pp.39-44
• /
• 2003

As one of many design variables, the role of dimension tolerances is to restrict the amount of size variation in a manufactured feature while ensuring functionality. In this study, a nonlinear integer model has been modeled to allocate the optimal tolerance to each individual feature at a minimum manufacturing cost. While a normal distribution determines statistically worst tolerances with its symmetrical property in many previous tolerance allocation studies, a asymmetrical distribution is more realistic because its mean is not always coincident with a process center. A nonlinear integer model is modeled to allocate the optimal tolerance to a feature based on a beta distribution at a minimum total cost. The total cost as a function of tolerances is defined by machining cost and quality loss. After the convexity of manufacturing cost is checked by the Hessian matrix, the model is solved by the Complex Method. Finally, a numerical example is presented demonstrating successful model implementation for a nonlinear design case.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.11 no.4
• /
• pp.39-46
• /
• 2007

One of well known and much studied nonlinear matrix equations is the matrix polynomial which has the form G(X)=$A_0X^m+A_1X^{m-1}+{\cdots}+A_m$ where $A_0$, $A_1$, ${\cdots}$, $A_m$ and X are $n{\times}n$ real matrices. We show how the minimization methods can be used to solve the matrix polynomial G(X) and give some numerical experiments. We also compare Polak and Ribi$\acute{e}$re version and Fletcher and Reeves version of conjugate gradient method.

• Communications of the Korean Mathematical Society
• /
• v.36 no.4
• /
• pp.783-800
• /
• 2021

We prove the non-existence of warped product GCR-lightlike submanifolds of the type K_⊥ × _λ K_T such that K_T is a holomorphic submanifold and K_⊥ is a totally real submanifold in an indefinite Kaehler manifold $\tilde{K}$. Further, the existence of GCR-lightlike warped product submanifolds of the type K_T × _λ K_⊥ is obtained by establishing a characterization theorem in terms of the shape operator and the warping function in an indefinite Kaehler manifold. Consequently, we find some necessary and sufficient conditions for an isometrically immersed GCR-lightlike submanifold in an indefinite Kaehler manifold to be a GCR-lightlike warped product, in terms of the canonical structures f and ω. Moreover, we also derive a geometric estimate for the second fundamental form of GCR-lightlike warped product submanifolds, in terms of the Hessian of the warping function λ.

• Bulletin of the Korean Chemical Society
• /
• v.10 no.3
• /
• pp.273-278
• /
• 1989

The gas-phase reactions of methyl formate with anions, $-NH_2,\;-OH,\;-CH_2CN$, are studied theoretically using the AM1 method. Stationary points are located by the reaction coordinate method, refined by the gradient norm minimization and characterized by the determination of Hessian matrix. Potential energy profiles and the stationary point structures are presented for all conceivable processes. Four reaction paths are found to be possible: formyl proton and methyl proton abstractions, carbonyl addition, and $S_N2$ process. For the most basic anion $-NH_2$ the proton abstraction path is favored, while in other case, $OH\;and\;-CH_2CN$, the carbonyl addition paths are favored. In all cases the $S_N2$ process is the most exothermic, but due to the relatively high activation barrier the process can be ruled out.

• International Journal of Computer Science & Network Security
• /
• v.21 no.6
• /
• pp.200-206
• /
• 2021

Fraud in e-commerce transaction increased in the last decade especially with the increasing number of online stores and the lockdown that forced more people to pay for services and groceries online using their credit card. Several machine learning methods were proposed to detect fraudulent transaction. Neural networks showed promising results, but it has some few drawbacks that can be overcome using optimization methods. There are two categories of learning optimization methods, first-order methods which utilizes gradient information to construct the next training iteration whereas, and second-order methods which derivatives use Hessian to calculate the iteration based on the optimization trajectory. There also some training refinements procedures that aims to potentially enhance the original accuracy while possibly reduce the model size. This paper investigate the performance of several NN models in detecting fraud in e-commerce transaction. The backpropagation model which is classified as first learning algorithm achieved the best accuracy 96% among all the models.

• Coupled systems mechanics
• /
• v.11 no.2
• /
• pp.133-149
• /
• 2022

The accurate prediction of elastoplasticity under prescribed workloads is essential in the optimization of engineering structures. Mechanical experiments are carried out with the goal of obtaining reliable sets of material parameters for a chosen constitutive law via inverse identification. In this work, two sample geometries made of high strength steel plates were evaluated to determine the optimal configuration for the identification of Ludwik's nonlinear isotropic hardening law. Finite element model updating(FEMU) was used to calibrate the material parameters. FEMU computes the parameter changes based on the Hessian matrix, and the sensitivity fields that report changes of computed fields with respect to material parameter changes. A sensitivity analysis was performed to determine the influence of the sample geometry on parameter identifiability. It was concluded that the sample with thinned gauge region with a large curvature radius provided more reliable material parameters.

• Journal of Advanced Information Technology and Convergence
• /
• v.9 no.1
• /
• pp.89-102
• /
• 2019

Collagen can be used in building artificial skin replacements for treatment of burns and towards the reconstruction of bone as well as researching cell behavior and cellular interaction. The strength of collagen in connective tissue rests on the characteristics of collagen fibers. 3D confocal imaging of collagen fibers enables the characterization of their spatial distribution as related to their function. However, the image stacks acquired with confocal laser-scanning microscope does not clearly show the collagen architecture in 3D. Therefore, we developed a new method to reconstruct, visualize and characterize collagen fibers from fluorescence confocal images. First, we exploit the tensor voting framework to extract sparse reliable information about collagen structure in a 3D image and therefore denoise and filter the acquired image stack. We then propose to segment the collagen fibers by defining an energy term based on the Hessian matrix. This energy term is minimized by a min cut-max flow algorithm that allows adaptive regularization. We demonstrate the efficacy of our methods by visualizing reconstructed collagen from specific 3D image stack.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.300-302
• /
• 2022

Palm vein identification has attracted attention due to its distinct characteristics and excellent recognition accuracy. However, many contactless palm vein identification systems suffer from the issue of having low-quality palm images, resulting in degradation of recognition accuracy. This paper proposes the use of U-Net architecture to correctly segment the vascular blood vessel from palm images. Attention gate mechanism and residual block are also utilized to effectively learn the crucial features of a specific segmentation task. The experiments were conducted on CASIA dataset. Hessian-based Jerman filtering method is applied to label the palm vein patterns from the original images, then the network is trained to segment the palm vein features from the background noise. The proposed method has obtained 96.24 IoU coefficient and 98.09 dice coefficient.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.6
• /
• pp.48-60
• /
• 2011

Nowadays, the 3D community is actively researching on 3D imaging and free-viewpoint video (FVV). The free-viewpoint rendering in multi-view video, virtually move through the scenes in order to create different viewpoints, has become a popular topic in 3D research that can lead to various applications. However, there are restrictions of cost-effectiveness and occupying large bandwidth in video transmission. An alternative to solve this problem is to generate virtual views using a single texture image and a corresponding depth image. A critical issue on generating virtual views is that the regions occluded by the foreground (FG) objects in the original views may become visible in the synthesized views. Filling this disocclusions (holes) in a visually plausible manner determines the quality of synthesis results. In this paper, a new approach for handling disocclusions using depth based inpainting algorithm in synthesized views is presented. Patch based non-parametric texture synthesis which shows excellent performance has two critical elements: determining where to fill first and determining what patch to be copied. In this work, a noise-robust filling priority using the structure tensor of Hessian matrix is proposed. Moreover, a patch matching algorithm excluding foreground region using depth map and considering epipolar line is proposed. Superiority of the proposed method over the existing methods is proved by comparing the experimental results.