• Proceedings of the Korean Magnestics Society Conference
• /
• 2011.12a
• /
• pp.176-177
• /
• 2011

• Proceedings of the Korean Vacuum Society Conference
• /
• 2007.08a
• /
• pp.166-166
• /
• 2007

• Proceedings of the Korean Nuclear Society Conference
• /
• 1999.10a
• /
• pp.194-194
• /
• 1999

• Proceedings of the Korean Radioactive Waste Society Conference
• /
• 2016.10a
• /
• pp.249-250
• /
• 2016

• Communications for Statistical Applications and Methods
• /
• v.4 no.2
• /
• pp.327-332
• /
• 1997

다층 신경망은 비모수 회귀함수 추정의 한 방법이다. 다충 신경망을 학습시키기 위해 역전파 알고리즘이 널리 사용되고 있다. 그러나 이 알고리즘은 이상치에 매우 민감하여 이상치를 포함하고 있는 자료에 대하여 원하지 않는 회귀함수를 추정한다. 본 논문에서는 통계물리에서 자주 사용하는 방법을 이용하여 로버스트 역전파 알고리즘을 제안하고 수학적으로 신경망과 매우 유사한 PRP(projection pursuit regression) 방법, 일반적인 역전파 알고리즘과 모의실험을 통해 비교 분석한다.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2007.11a
• /
• pp.71.2-71.2
• /
• 2007

• The Journal of Korean Association of Computer Education
• /
• v.20 no.6
• /
• pp.83-94
• /
• 2017

The purpose of this study is to analyze factors affecting elementary school students' computational thinking and SW education attitude using a 3-level multi-level models. The results of this study are as follows: First, 'Computer at home', 'SW competition participation experience', 'SW education satisfaction', and 'SW awareness' have a statistically significant effect on the initial value of computational thinking while 'SW period of education' and 'SW education experience at after school' have a statistically effect on the change rate of computational thinking. Second, 'SW awareness', 'SW education satisfaction' and 'gender' have a positive effect on the initial values of SW education attitude whereas 'SW period of education' has a slight negative influence on the change rate of SW education attitude.

• Journal of KIISE:Software and Applications
• /
• v.26 no.12
• /
• pp.1383-1392
• /
• 1999

이미 학습된 다층퍼셉트론 신경망을 디지털 VLSI 기술을 사용하여 하드웨어로 구현할 경우 신경망의 가중치 및 뉴런 출력들을 양자화해야 하는 문제가 발생한다. 이러한 신경망 변수들의 양자화는 결과적으로 주어진 입력에 대한 신경망의 최종 출력에서의 왜곡을 초래한다. 본 논문에서는 먼저 이러한 양자화로 인한 신경망 출력에서의 왜곡을 통계적으로 분석하였다. 분석 결과에 의하면 입력패턴 각 성분의 제곱들의 합과 가중치의 크기들이 양자화 영향에 주로 기여하는 것으로 나타났다. 이러한 분석 결과를 이용하여 양자화를 위한 정밀도가 주어졌을 때, 양자화 영향이 최소화된 다층퍼셉트론 신경망을 설계하는 방법을 제시하였다. 그리고 제안된 방법에 의해 얻은 신경망과 오류역전파 학습방법에 의하여 얻은 신경망의 성능을 비교함으로써 제안된 방법의 효율성을 입증하였다. 실험결과는 낮은 양자화 정밀도에서도 제안된 방법이 더 좋은 성능을 보였다.Abstract When we implement a multilayer perceptron with the digital VLSI technology, we generally have to quantize the weights and the neuron outputs. These quantizations eventually cause distortion in the output of the network for a given input. In this paper first we made a statistical analysis about the effect caused by the quantization on the output of the network. The analysis revealed that the sum of the squared input components and the sizes of the weights are the major factors which contribute to the quantization effect. We present a design method for an MLP which minimizes the quantization effect when the precision of the quantization is given. In order to show the effectiveness of the proposed method, we developed a network by our method and compared it with the one developed by the regular backpropagation. We could confirm that the network developed by our method performs better even with a low precision of the quantization.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.3
• /
• pp.209-214
• /
• 2017

A plane-wave incident upon asymmetric multi-layered dielectric grating as well as symmetric grating structure generates space harmonics. Selected space harmonics among those harmonics can undergo strong resonance scattering variations known as GMR(guided-mode resonance). In this paper, to clarify these effects, the field propagation and dispersion curve inside the grating region are analyzed by using a rigorous equivalent transmission-line theory(RETT) based on eigenvalue problem. The results show that, at the peak of a scattering resonance, the reflected mode is almost identical to a leaky wave that can be supported by the grating structure. Thus, it confirms to be occurred GMR effect associated with the free-resonant character of leaky waves at asymmetric multi-layered dielectric gratings. Quantitative simulation results illustrating the behavior of typical gratings are given, and the special case of normal incidence is discussed for TE and TM modes.

• Korean Journal of Optics and Photonics
• /
• v.6 no.4
• /
• pp.259-265
• /
• 1995

Antireflection coatings on optical lenses commertially available in domestic market are optically analyzed. Transmission spectra and reflection spectra are collected using spectrophotometers. The apparent absorption spectra around the absorption band edge are dominated by the substrate absorption. The reflection spectra and the apparent absorption spectra at visible region between 400nm and 700nm show very strong correlation to each other except a couple samples. The discrepency observed in the latters are due to an increased absorption in visible region by the substrate, which is negative effect of these samples. An antireflection coating consisted of $SiO_2/TiO_2/SiO_2/ZrO_2/Cr$ is made on c-Si substrate for spectroscopic ellispometry analysis. A film-by-film coating is accomplished and between each film deposition, ex-situ spectroscopic ellipsometry measurements are made. The analysis of the spectroscopic ellipsometry data reveals that the average film densities of $ZrO_2$ and $TiO_2$ reach only 80% of their respective packing densities and thick films are inhomogeneous along film growth direction. Discussions are made toward in-situ, real-time monitoring of the film growth so that a real-time feedback is possible to achieve a post-correction to minor deviations occured in the previous step. step.