• The KIPS Transactions:PartB
• /
• v.12B no.3 s.99
• /
• pp.343-348
• /
• 2005

The goal of this paper is to study the joint effect of factors of neural network teaming procedure. There are many factors, which may affect the generalization ability and teaming speed of neural networks, such as the initial values of weights, the learning rates, and the regularization coefficients. We will apply a constructive training algerian for neural network, then patterns are trained incrementally by considering them one by one. First, we will investigate the effect of these factors on generalization performance and learning speed. Based on these factors' effect, we will propose a joint method that simultaneously considers these three factors, and dynamically hue the learning rate and regularization coefficient. Then we will present the results of some experimental comparison among these kinds of methods in several simulated nonlinear data. Finally, we will draw conclusions and make plan for future work.

• Journal of KIISE
• /
• v.43 no.7
• /
• pp.812-819
• /
• 2016

This paper introduces and analyzes the theoretical basis and method of the conventional initial-value assignment problem and feasibility of image segmentation. The paper presents topological evidence and a method of appropriate initial-value assignment based on topology theory. Subsequently, the paper shows minimum conditions for feasibility of image segmentation based on separation axiom theory of topology and a validation method of effectiveness for image modeling. As a summary, this paper shows image segmentation with its mathematical validity based on topological analysis rather than statistical analysis. Finally, the paper applies the theory and methods to conventional Gaussian random field model and examines effectiveness of GRF modeling.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.28 no.1B
• /
• pp.111-114
• /
• 2008

Explicit solutions of the wave dispersion equation are developed using the recursive relation in terms of the relative water depth. We use the solutions of Eckart (1951), Hunt (1979), and the deep-water and shallow-water solutions for initial values of the solution. All the recursive solutions converge to the exact one except that with the initial value of deep-water solution. The solution with the initial value by Hunt converged much faster than the others. The recursive solutions may be obtained quickly and simply by a hand calculator. For the transformation of linear water waves in whole water depth, the use of the recursive solutions will yield more accurate analytical solutions than use of previously developed explicit solutions.

• Journal of the Korean Society for Railway
• /
• v.15 no.3
• /
• pp.278-285
• /
• 2012

The Service Availability (SA) in the viewpoint of passenger is used as the key performance indicator (KPI) of quality of service in Light Rail Transit (LRT) Public-Private Partnerships projects. But there are many disputes on the target value of SA because of the lack of experience in SA. The target value of SA should be set at an early stage of the project to be specified on the system specifications and operation plan. Therefore, this paper developed the quantitative prediction model of SA to set the reasonably achievable target value of SA at an early stage of the LRT project. Also this paper analyzed the relationship and differentiation of SA and Train Punctuality (TP) that is mostly compared with SA.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.05a
• /
• pp.19-24
• /
• 2003

신경망은 적용 다양성과 제약조건의 최소성, 강력한 예측성, 범용성, 근사성 등 많은 장점을 지니고 있으나 초기 가중치의 할당에 따라 모델 생성의 Performance와 예측의 결과가 달라지게 되는 단점을 지니고 있다. 이런 신경망의 초기 가중치에 따른 단점을 보안하기 위해 통계적 알고리즘의 접목을 통해 Hybrid된 신경망 보완 알고리즘을 제시하고자 하였다. 논문을 위한 기본 가정으로 신경망의 가장 기본인 SLP 알고리즘을 바탕으로 활성함수에 가장 일반적으로 사용되는 Sigmoid 활성함수를 이용하였을 때, 초기 가중치로 기존의 임의 난수 생성 방식이 아닌 통계적 로지스틱 회귀분석의 계수값(mle)을 제시하여 이를 초기치로 사용한 경우와 그렇지 않은 경우의 예측 정확성과 수렴의 Performance정도를 비교하여 가장 효과적인 초기치 방법을 제시하고자 하였다.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.35 no.5
• /
• pp.622-628
• /
• 2006

Kochujang was prepared with various concentrations of Maesil extracts (0, 1, 2, and 5%) and the physicochemical characteristics of Maesil extracts added Kochujang ('Maesil Kochujang') were investigated during aging of 100 days. pH decreased slightly during aging. Titratable acidity, on the other hand, increased with aging, reaching at the highest level at the 40 days of aging and then decreased slowly. The changes of moisture content were complicated in the beginning, but after 60 days it increased slowly. The water activity decreased but salt concentration increased consistently in all samples during aging. L* and a*-values decreased during aging and L*, a*, and b*-values decreased significantly with high amount of Maesil extract in the sample. Soluble solid contents increased slowly while reducing sugar contents increased for up to 40 days and then decreased. Amino nitrogen contents increased steadily after 40 days of aging and they reached at $230.00{\sim}246.00mg%$ upon 100 days of aging.

• Progress in Medical Physics
• /
• v.26 no.4
• /
• pp.229-233
• /
• 2015

We retrospectively reviewed lung cancer patients who were treated with stereotactic ablative radiotherapy (SABR). We investigated the value of response evaluation after treatment by measuring the volume change of tumors on serial chest computed tomography (CT) examinations. The study included 11 consecutive patients with early-stage (T1-T2aN0M0) non-small cell lung cancer (NSCLC) who were treated with SABR. The median dose of SABR was 6,000 cGy (range 5,000~6,400) in five fractions. Sequential follow-up was performed with chest CT scans. Median follow-up time was 28 months. Radiologic measurement was performed on 51 CT scans with a median of 3 CT scans per patient. The median time to partial response ($T_{PR}$) was 3 months and median time to complete remission ($T_{CR}$) was 5 months. Overall response rate was 90.9% (10/11). Five patients had complete remission, five had partial response, and one patient developed progressive disease without response. On follow-up, three patients (27.2%) developed progressive disease after treatment. We evaluated the the response after SABR. Our data also showed the timing of response after SABR.

• The Bulletin of The Korean Astronomical Society
• /
• v.37 no.1
• /
• pp.70.1-70.1
• /
• 2012

CQUEAN (Camera for QUasars in EArly uNiverse)은 초기우주천체 연구단(Center for Exploration of Origin of the Universe) 사업에서 개발한 CCD 카메라로서 초기우주의 퀘이사 후보를 찾기 위한 목적으로 설계되었다. CCD를 구동할 때는 픽셀 다이오드의 PN 접합층에 공핍층(depletion layer)을 생성하기위해 역 바이어스 전압을 준다. 이 전압에 의해 CCD를 사용한 관측 시 광이온화와 열이온화 현상에 의해 생성된 전자의 전하값에 추가로 바이어스 값이 읽혀진다. 정확한 CCD 측광 결과를 얻어내기 위해서는 안정된 바이어스를 유지해야 한다. 본 연구에서는 향후 CQUEAN의 보다 정확한 관측 및 데이터 처리에 대비하여 CQUEAN의 바이어스 특성을 분석하여 이 값에 영향을 주는 요인을 찾고 해결책을 논의한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.839-841
• /
• 2013

LTE 환경에서 초기 망 접속을 위한 UE와 MME 간의 사용자의 정보에 대한 인증을 위하여 IMSI, GUTI 등의 정보를 요청, 승인 하게 된다. 이 과정에서 사용자의 유일한 식별값인 IMSI를 새로운 MME로 전송하는 경우 별도의 암호화 과정 없이 평문의 IMSI 값이 전송되어진다. 이로 인하여 평문으로 전송되는 IMSI값에 대한 보안 위협이 발생하고, 이로 인하여 사용자 추적, 단말 추적 등 사용자의 Privacy에 대한 침해 및 가장 공격 등이 발생된다. 본 논문에서는 UE와 MME간 초기 통신과정에서 IMSI에 대한 암호화를 통하여 안전성 강화에 대한 제안을 한다.

• Geophysics and Geophysical Exploration
• /
• v.1 no.1
• /
• pp.49-56
• /
• 1998

Seismic tomography has been widely used as high resolution subsurface imaging techniques in engineering applications. Although most of the techniques have been using travel time inversion, waveform method is being driven forward owing to the progress of computational environments. Although full-waveform inversion method has been known as the best method in terms of model resolving power without high-frequency restriction and weak scattering approximation, it has practical disadvantage that it is apt to get stuck in local minimum if the initial guess is far from the actual model and it consumes so much time to calculate. In this study, 2-D full-waveform inversion algorithm in acoustic medium is developed, which uses result of traveltime tomography as initial model. From the application on synthetic data, it is proved that this approach can efficiently reduce the problem of conventional approaches: our algorithm shows much faster convergence rate and improvement of model resolution. Result of application on physical modeling data also shows much improvement. It is expected that this algorithm can be applicable to real data.