• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.10
• /
• pp.73-81
• /
• 2015

The paper proposes a prototype selection method and evaluates the generalization performance of standard algorithms and prototype based classification learning. The proposed prototype classifier defines multidimensional spheres with variable radii within class areas and generates a small set of training data. The nearest-neighbor classifier uses the new training set for predicting the class of test data. By decomposing bias and variance of the mean expected error value, we compare the generalization errors of k-nearest neighbor, Bayesian classifier, prototype selection using fixed radius and the proposed prototype selection method. In experiments, the bias-variance changing trends of the proposed prototype classifier are similar to those of nearest neighbor classifiers with all training data and the prototype selection rates are under 27.0% on average.

• Journal of KIISE
• /
• v.45 no.3
• /
• pp.280-287
• /
• 2018

Recently, the use of a recommendation system and tensor data analysis, which has high-dimensional data, is increasing, as they allow us to analyze the tensor and extract potential elements and patterns. However, due to the large size and complexity of the tensor, it needs to be decomposed in order to analyze the tensor data. While several tools are used for tensor decomposition such as rTensor, pyTensor, and MATLAB, since such tools run on a single machine, they are unable to handle large data. Also, while distributed tensor decomposition tools based on Hadoop can handle a scalable tensor, its computing speed is too slow. In this paper, we propose S-PARAFAC, which is a tensor decomposition tool based on Apache Spark, in distributed in-memory environments. We converted the PARAFAC algorithm into an Apache Spark version that enables rapid processing of tensor data. We also compared the performance of the Hadoop based tensor tool and S-PARAFAC. The result showed that S-PARAFAC is approximately 4~25 times faster than the Hadoop based tensor tool.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2016.02a
• /
• pp.113-113
• /
• 2016

반도체 산업이 발전하고 기술이 향상됨에 따라 미세화되고 복잡한 구조의 소자가 개발되고 있으며, 2차원 소재 등 다양하고 새로운 소재들이 발견 및 연구되고 있다. 새로운 소재 또는 기술을 이용한 고품질 소자를 개발하기 위해서는 우수한 특성(높은 순도, 우수한 분해 및 반응 특성)을 지닌 증착소재의 개발 및 평가가 선행되어야 한다. 기존의 증착소재의 기본 물성을 측정하는 방법인 단순 기상 Fourier transform infrared spectroscopy(FT-IR) 분석법은 실제 공정에서의 증착경향을 대변하기 어렵다는 단점이 있다. 이러한 단점을 보완하기 위해 개조된 attenuated total reflection (ATR) 액세서리를 이용하여 실제 공정에서의 증착경향을 대변하고자 하였다. 본 연구에서는 반도체 증착소재의 분해 및 표면 흡착 특성을 분석하기 위해 ATR-FTIR 분석법을 이용하여 수행하였으며, 분산안정도에 따른 nanoparticle을 ATR의 크리스탈 표면에 분포시켜 hexamethyldisilazane(HMDS) source의 흡착 효율을 향상시키는 연구를 수행하였다. Nanoparticle의 분산안정도를 높이기 위하여 suspension 상태에서 pH, sonication, 분산제를 이용하였으며, nanoparticle을 ATR crystal 표면에 분포하여 분석한 결과, 분산안정도에 따라 HMDS의 흡착효율이 달라짐을 확인하였다.

• International Area Studies Review
• /
• v.13 no.3
• /
• pp.309-330
• /
• 2009

This paper analyzes comparatively the effects of changes in real effective exchange rates on the business fluctuations of the cases of Korea and Japan employing structural vector auto-regression(S-VAR) model which uses quarterly data for the five variables of real effective exchange rates, GDP gap, real interest rates, oil prices, inflation rates for the period of 1980-2006. The paper employes impulse-response analysis and variance decompositions. The paper finds that real exchange rate depreciations are contractionay for the case of Korea while they are expansionary for the case of Japan. These results are consistent with the prevailing empirical results that real exchange rate depreciations are contractionary for developing countries while expansionary for advanced countries.

• Textile Coloration and Finishing
• /
• v.8 no.5
• /
• pp.25-48
• /
• 1996

이 연구의 목적은 에틸렌이민 유도체를 갖는 아조계 분산 염료의 폴리에스테르 섬유에의 응용과 최적화된 폴리에스테르 염색 조건의 확립이다. 이들 아지리디닐 염료는 구조가 유사하며 말단에 하이드록시기를 갖는 염료(아지리디닐 염료의 가수분해 형태)와 비교하여 최대 흡수 파장의 천색적으로 천이 됨을 알 수 있었다. 세가지 아지리디닐 모노 아조 염료와 이들 염료의 가수분해 된 형태 그리고 디메틸아니린계 아조 염료의 폴리에스테르 섬유에 대한 염색 및 견뢰도 성질이 조사되었다. 아지리디닐 아조 염료는 산성 조건에서보다 염기성 조건하에서 보다 안정하였고, 뛰어난 build-up이 높은 pH조건에서 얻어졌으며, 폴리에스테르 염색 시 130$^{\circ}C$, pH 10의 조건에서 최적화된 색상 강도를 얻었다. 아지리디닐 염료는 이 염료의 가수분해 형태와 디메틸아닐린 염료와 비교하여 개선된 세탁 견뢰도를나타내었고 이는 용매 추출을 이용하여 확인 하였다. 아지리디닐 아조 염료는 역시 광 련뢰도에서도 또한 상용화된 분산 염료에 비교하여 높은 결과를 나타내었다.

• Management & Information Systems Review
• /
• v.31 no.4
• /
• pp.287-307
• /
• 2012

We empirically examine the price discovery dynamics among the VKOSPI, the KOSPI200 spot, and the KOSPI200 futures markets. The analysis employs the vector-autoregression, Granger causality, impulse response function, and variance decomposition using both daily data from 2009. 04. 13 to 2011. 12. 30 and 1 minute data from the bull market, bear market, and the flat period. The main results are as follows; First, the lead lag relationships between KOSPI200 spot(futures) yield VKOSPI returns could not be found from the daily data analysis. But KOSPI200 spot(futures) have a predictive power for VKOSPI from 1 minute data. Especially KOSPI200 spot(futures) and VKOSPI show the bi-directional effects to each other during the return rising period Second, We chose the VAR(1) the model in daily data but adopt the VAR(3) model in the one minute data to determine the lead lag time. We know that there is predictability during the very short period Third, Spot returns and futures returns makes no difference in daily data results. According to the one minite data results, VKOSPI returns have a predictive power for KOSPI200 spot return, but have no predictive power for KOSPI200 futures return.

• Polymer(Korea)
• /
• v.32 no.3
• /
• pp.246-250
• /
• 2008

Colloidal stability of the biodegradable nanoparticle was characterized by measuring the variation of particle size with time using photon correlation spectroscopy. Three kinds of polymers, namely, poly(D,L-lactide-co-glycolide)(PLGA), PLGA/poly(L-lactide) blends, and PLGA/poly(L-lactide)-g-poly(ethylene glycol) blends were used as matrix material for nanoparticle preparation. Nanoparticles were prepared with or without using poly(vinyl alcohol)(PVA) as suspension stabilizer to evaluate the condition of preparation. Nanoparticles from the blend of amphiphilic graft copolymer with short poly(ethylene glycol) chain and PLGA maintained suspension for 1 day when protein stock solution was introduced. This is somewhat improvement in colloidal stability against protein adsorption compared with that of nanoparticles without PEG moiety. Suspension stabilizer, PVA, had a significant effect on the colloidal stability against freezing and protein adsorption which led to coagulation of nanoparticles. It is important to consider effect of suspension stabilizer as well as materials used to prepare nanoparticle on the colloidal stability.

• Applied Chemistry for Engineering
• /
• v.18 no.5
• /
• pp.490-494
• /
• 2007

Dispersants were prepared from petroleum pyrolysis residual oil (PPRO) through sulfonation. Without employing a conventional polynaphthalene sulfonate formaldehyde condensate (PNS) process, the dispersants (NPS) were synthesized by a simpler process only in 2 h. The chemical structure of new dispersant, which has various naphthalene derivative groups, was similar to PNS conformed by UV-visible spectroscopy curves. The new dispersants demonstrated high dispersing ability in inorganic suspension ; cement, $Fe_2O_3$, and $CaCO_3$.

• Proceedings of the Safety Management and Science Conference
• /
• 2011.04a
• /
• pp.571-572
• /
• 2011

이 논문은 주식시장에서 유동성이 감소될 경우 단기에 있어서 주가수익률의 하락현상에 대한 상관관계와 관련된 것이다. 그리고 주식시장에서의 유동성 감소가 기업들의 수익성 악화로 연결될 수 있는지와 관련된 연구를 하였다. 분석결과에 따르면, 한국에 있어서 Granger 인과성 검정결과와 충격반응분석, 분산분해 분석결과에 따르면 채산성 증감률의 경우 코스피수익률보다 코스닥수익률에 더 영향이 큼을 알 수 있었는데, 이는 소형주가 자금사정에 더 민감함을 반영하고 시차도 길게 반영되고 있음을 나타내고 있다. 이는 주식시장에서의 유동성 감소가 기업들의 수익성 악화로 연결되는데, 대형주에 비해서는 소형주에 보다 영향을 크게 미칠 수 있음을 시사하고 있다.

• Proceedings of the Korean Fiber Society Conference
• /
• 1992.06b
• /
• pp.45-45
• /
• 1992