• 응용통계연구
• /
• 제25권1호
• /
• pp.225-237
• /
• 2012

This study proposes a quantification algorithm for a PLS method with several sets of variables. We called the quantification method for PLS with more than 2 sets of data a generalization. The basis of the quantification for PLS method is singular value decomposition. To derive the form of singular value decomposition in the data with more than 2 sets more easily, we used the constraint, $a^ta+b^tb+c^tc=3$ not $a^ta=1$, $b^tb=1$, and $c^tc=1$, for instance, in the case of 3 data sets. However, to prove that there is no difference, we showed it by the use of 2 data sets case because it is very complicate to prove with 3 data sets. The keys of the study are how to form the singular value decomposition and how to get the coordinates for the plots of variables and observations.

• Communications for Statistical Applications and Methods
• /
• 제16권3호
• /
• pp.421-428
• /
• 2009

범주형 자료의 시각화에서 범주가 많지 않은 경우에는 기존의 Hayashi의 수량화 제3방법을 이용하여 두변수의 범주들 사이의 연관성에 대한 시각화를 구할 수 있다. 그러나, Hayashi방법은 큰 빈도의 범주들보다 작은 빈도의 범주들을 두드러지게 수량화하므로 결과가 불안정하다는 문제점이 있다 (허명회와 이용구, 2006). 이 연구의 목적은 범주수가 "큰" 두 범주형 변수 R과 C에 대하여 각 변수 벌주들 사이의 연관성을 살펴보기 위한 시각화 방법을 제안하는 데 있다. 이를 위하여 우리는 2개 변수군 수치형 자료를 시각화하는 방법으로 제안된 허명회 등 (2007)의 PLS 시각화 방법을 범주형 자료에 적용하고자 한다. 즉, 범주형 변수 R과 C의 범주들 각각을 0/1로 더미 코드화하여 각각 R개와 C개의 범주군으로 변환한 다음 허명회 등 (2007)에서 제시한 PLS 시각화 방법을 적용하고자 한다. 이러한 방법은 Hayashi 수량화 방법의 문제점을 해결할 수 있을 뿐만 아니라 행변수와 열변수 각각이 여러 개의 범주형 변수들의 집합인 변수군의 경우에도 확대 적용 가능하다. 순치 예로서 German Credit 자료에서 10개 금융관련 변수의 34개 범주를 R로 간주하고 10개 사회인구적 변수의 46개 범주를 C로 간주하여 새 방법론을 적용해 보인다.

• 응용통계연구
• /
• 제20권2호
• /
• pp.345-355
• /
• 2007

PLS 회귀는 q-변량의 Y 변수에 대한 회귀에서 p-변량의 X 변수가 다중공선성의 문제를 갖는 경우에도 적용 가능한 방법이다. 특히 X 변수의 수 p가 관측개체 수 n보다 큰 경우에 적용 가능하여 계량화학(chemometrics) 분야에서 근적외선 분광기(near-infrared spectroscopy) 자료에 대한 표준적 분석 방법으로 활용되고 있다. 이 연구에서 우리는 PLS회귀의 방법론을 정리하고 이를 활용한 p개의 X 변수들과 q개의 Y 변수들의 동시 시각화를 위한 두 가지의 수량화 방법을 제안한다.

• 품질경영학회지
• /
• 제30권3호
• /
• pp.38-52
• /
• 2002

This paper is concerned with the analysis of structural equation model(SEM) with the ordinal data such as Likert scale. The SEM is misused when the arbitrary scores allocated to the Likert scale are treated as quantitative data. The underlying distribution approaches have been studied to solve this problem, and the partial least squares(PLS) Is also tried. In this paper the quantification methods for the Likert scale are proposed to analyze the SEM. We assume that the Likert scale is an observation of the interval of the continuous underlying distribution, and the respondents have their own patterns in the response of some questions. Normal and beta distributions as the response patterns are considered to quantify the Likert scale. To compare the efficiency of the proposed method the bootstrap simulations are tried.

• 한국근적외분광분석학회:학술대회논문집
• /
• 한국근적외분광분석학회 2001년도 NIR-2001
• /
• pp.4111-4111
• /
• 2001

When a drug is prepared in a tablet, the active component represents only a small portion of the dosage form. The other components of the formulation include materials to assist in the dissolution, antioxidants, coloring agents and bulk fillers. The tablets are tested using approved testing methods usually involving separation and subsequent quantification of the active component. Tablets may also be tested by near-Infrared Reflectance spectrometry (NIRS). In the present study, based on NIRS and multivariate calibration methods, a novel and precise method is developed for direct determination of ascorbic acid in vitamin C tablet. Two different tablet formulations were powdered in three different sizes, 63-125 ${\mu}{\textrm}{m}$, and examined. Spectral region of 4750-4950 $cm^{-1}$ / was used and optimized for quantitative operations. Partial least squares (PLS) and multiple linear regression (MLR) methods were performed for this spectral region. The results of optimized PLS and MLR methods showed that reproducibility increase with decreasing grain size and standard error of calibration (SEP) of less than 1% w/w of ascorbic acid and a correlation coefficient of 0.998 can be achieved. The PLS method showed better results than MLR. Seven overdose and underdose samples (prepared in the laboratory to match marketed products) were tested by proposed and iodometric standard methods. A correlation between NIRS predicted ascorbic acid values and iodomet.ic values was calculated ($R^2$=0.9950). Finally, the direct analysis of individual intact tablets in their unit-dose packages (Blistering in aluminum and PVC foils) obtained from market were also carried out and a correlation coefficient of 0.9989 and SEP of 0.931% w/w of ascorbic acid were achieved.

• 대한원격탐사학회지
• /
• 제37권4호
• /
• pp.777-788
• /
• 2021

Since aerosol has a relatively short duration and significant spatial variation, satellite observations become more important for the spatially and temporally continuous quantification of aerosol. However, optical remote sensing has the disadvantage that it cannot detect AOD (Aerosol Optical Depth) for the regions covered by clouds or the regions with extremely high concentrations. Such missing values can increase the data uncertainty in the analyses of the Earth's environment. This paper presents a spatial gap-filling framework using a univariate statistical method such as DCT-PLS (Discrete Cosine Transform-based Penalized Least Square Regression) and FMM (Fast Matching Method) inpainting. We conducted a feasibility test for the hourly AOD product from AHI (Advanced Himawari Imager) between January 1 and December 31, 2019, and compared the accuracy statistics of the two spatial gap-filling methods. When the null-pixel area is not very large (null-pixel ratio < 0.6), the validation statistics of DCT-PLS and FMM techniques showed high accuracy of CC=0.988 (MAE=0.020) and CC=0.980 (MAE=0.028), respectively. Together with the AI-based gap-filling method using extra explanatory variables, the DCT-PLS and FMM techniques can be tested for the low-resolution images from the AMI (Advanced Meteorological Imager) of GK2A (Geostationary Korea Multi-purpose Satellite 2A), GEMS (Geostationary Environment Monitoring Spectrometer) and GOCI2 (Geostationary Ocean Color Imager) of GK2B (Geostationary Korea Multi-purpose Satellite 2B) and the high-resolution images from the CAS500 (Compact Advanced Satellite) series soon.

• 한국근적외분광분석학회:학술대회논문집
• /
• 한국근적외분광분석학회 2001년도 NIR-2001
• /
• pp.4114-4114
• /
• 2001

For the quality control of tablets several parameters have to be checked. The most important one is the content of an active ingredient which has to match a narrow range around the designated content. The only useful measurement mode is transmission which provides information of the complete tablet. A measurement in diffuse reflectance would register only the surface which is useless especially in case of a coated tablet. In this work tablets for a clinical study (placebo/verum studies) with very low concentrations of the active ingredient were measured. The concentration range was 0 to 6 mg with a total weight of the tablets of 105 mg, leading to a highest concentration of the active component of 5.7% by weight. Especially the spectroscopic distinction between the placebo and the low dosage forms with 0.25 and 0.5 mg active agent requires an extraordinarily accurate sampling technique. Using the VECTOR 22/N-T in transmission mode allows the collection of the information from the complete tablets. A quantitative PLS-model with transmission spectra from the tablets described above shows that the active substance can be predicted with a RMSECV (root mean square error of cross validation) of 0.04% absolute for this special application. The results are compared with those of measurements in diffuse reflectance using different accessories.

• Bulletin of the Korean Chemical Society
• /
• 제29권1호
• /
• pp.61-68
• /
• 2008

High initial (2 minutes after iv injection) brain-uptake of PET agents is required to deliver the agent to binding sites in brain tissue but, for quantification of the specific binding, relatively rapid washout of free and non-specifically bound PET agents from the brain (30 minutes after injection) also is required. In order to compare the physicochemical properties of the PET agents which are responsible for early brain-uptake and rapid washout, respectively, chemometric analysis on brain-uptake of PET agents was performed via a classical VolSurf approach. According to the PCA and PLS results, high 2-30 min brain-uptake ratio seems to be related to the large hydrophobic regions in the PET agents which are not confined to a particular surface.

• 약학회지
• /
• 제49권2호
• /
• pp.115-121
• /
• 2005

In this study, the performance of a portable NIR system and a FT NIR spectrometer were compared to determine water content of hairless mouse skin. The stratum corneum parts wer e separated from the epidermal tissues by trypsin solution. NIR diffuse reflectance spectra of hairless mouse skin were acquired using a fiber optic probe. In the near infrared, water molecules show two clear absorption bands at 1450 nm from first overtone of O-H stretching and 1940 nm from the combination involving O-H stretching and O-H deformation. It was found that the variations of O-H absorption band according to water content. Partial least squares regression (PLSR) was applied to develop a calibration model. The PLS model showed a good correlation between NIR predicted value and the absolute water content of separated hairless mouse skin, in vitro. For both the portable and the FT NIR spectrometer, These studies showed the possibility of a rapid and nondestructive skin moisture measurement using NIR spectroscopy. The portable NIR spectrometer with a photodiode arrays-microsensor could be more rapidly applied for the determination of water content with comparable accuracy with the performance of a FT spectrometer .

• Journal of Plant Biotechnology
• /
• 제42권1호
• /
• pp.60-70
• /
• 2015

본 연구의 목적은 적외선 분광스펙트럼 데이터를 이용하여 대두 종자내의 지방산 함량을 동시에 예측할 수 있는지 여부를 조사하기 위한 것이다. 총 153종의 대두(Glycine max Merrill) 종자로부터 적외선 분광스펙트럼 및 지방산의 함량을 기체크로마토그라피 분석을 통하여 확인하였다. 적외선 분광스펙트럼 조사결과 대두는 단백질이나 아미노산의 amide bond region ($1,700{\sim}1,500cm^{-1}$), 핵산이나 인지질의 phosphodiester groups ($1,500{\sim}1,300cm^{-1}$) 그리고 탄수화물 등 다당류의 sugar region ($1,200{\sim}1,000cm^{-1}$)에서 계통별로 큰 차이가 이루어짐을 알 수 있었다. 총 29라인의 대두 계통별 시료로부터 지방산 함량을 조사한 결과 총 지방산의 함량은 건조 시료 0.1 g 당 $185.57{\mu}g$에서 $325.9{\mu}g$으로 계통간에 차이가 있었음을 알 수 있었으며 평균 함량은 $244.48{\mu}g$이었다. PLS regression 분석을 이용하여 총 5개 지방산(팔미틱산, 스테아릭산, 올레익산, 리노레익산 그리고 리노레닉산) 함량 예측 calibration models의 실측 검증 결과, 팔미틱산($R^2=0.8002$), 올레익산($R^2=0.8909$) 그리고 리노레익산($R^2=0.815$)은 회귀분석 상관계수가 0.8 이상으로 정확도 높음을 알 수 있었다. 그러나 스테아릭산($R^2=0.4598$)과 리노레닉산($R^2=0.6868$)의 경우 상관계수가 0.7 이하로 상대적으로 예측정확도가 낮음을 알 수 있었다. 본 연구에서 확립된 기술은 지방산의 조성 변환을 통하여 새로운 대두 품종 개발을 위한 계통선발 과정에서 매우 효율적인 수단으로 활용이 가능할 것으로 사료된다. 더 나아가 본 기술은 대두는 물론 대두 유래 농산물이나 식품의 품질 검증 수단으로 활용이 가능할 것으로 기대된다.