• Proceedings of the Korean Statistical Society Conference
• /
• 2003.05a
• /
• pp.151-156
• /
• 2003

Variable selection algorithm for principal component analysis using penalized likelihood method is proposed. We will adopt a probabilistic principal component idea to utilize likelihood function for the problem and use HARD penalty function to force coefficients of any irrelevant variables for each component to zero. Consistency and sparsity of coefficient estimates will be provided with results of small simulated and illustrative real examples.

• Journal of Animal Science and Technology
• /
• v.44 no.6
• /
• pp.643-652
• /
• 2002

Data were consisted of the ten body measurements (withers height, rump height, body length, chest depth, chest width, rump width, rump length, thurls width, hipbone width and chest girth) of 642 steers (Korean cattle), which was entered in the National Beef Quality Contest hosted by the Korea Animal Improvement Association from 1997 to 2001. A principal components analysis was used to classify the body types of the steers, and estimate the correlations between carcass traits and principal components for the body measurements of the first, second, third and fourth period, respectively. The first principal component of body measurements at the first, second, third and fourth period accounted for 76.0%, 83.0%, 72.7% and 57.4% of the total variance, respectively. The sum of first, second and third principal component at each period accounted for 86.69%, 90.49%, 84.62% and 77.26% of the total variance, respectively. At each period, all the first principal component of the body measurements were positive and it generally showed large framed body shape. The size of body was influenced mostly by chest depth(0.328${\sim}$0.339) and rump length(0.325${\sim}$0.341). The second, third and fourth principal component at the each period were various. There were positive correlations between principal components index of each period and carcass traits such as carcass weight(0.539${\sim}$0.755), average daily gain(0.256${\sim}$0.564), backfat thickness(0.227${\sim}$0.280), and eye muscle area(0.187${\sim}$0.344). The correlation with yield grade index(－0.246${\sim}$－0.110), however, was negative. The correlation with marbling score(0.066${\sim}$0.099) was low or statistically insignificant. According to principal component indexes of the second, third, and fourth components, the correlations with the carcass traits were various. There were no large differences between the correlations of the single body measurement trait with the carcass traits and the correlations of the first principal component indexes with the carcass traits.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.43 no.1
• /
• pp.6-10
• /
• 1998

This study was conducted to obtain basic information on the development of new triticale cultivars with good quality and high productivity for soiling feed. Twelve cultivars introduced from Poland, Canada and two cultivars developed in Korea were planted in the experimental field at Ansong National University in 1995. Major growth traits and nutrient components for feed were measured and analyzed using principal component analysis and average linkage cluster analysis. 'Prego', 'Prag 46/3', and 'Clercal' were relatively high in forage yield. Most of forage nutrient contents except cellulose were higher in Prego, Clercal, and 'Cumulus' than other cultivars. Results of principal component analysis on 11 traits including forage yield and nutrient contents showed that 72.59% of total variation were explained by the first and second principal components. The Z$_1$ had high correlation with the contents of forage nutrient components and Z$_2$ with plant height, fresh, and dry weight. Fourteen cultivars were classified into 7 groups by multivariate analysis. Clercal and Prego in Group I could be useful source for the improvement of triticale as an important forage crop because they exhibited high productivity as well as high contents of nutrient components for feed.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.35 no.1
• /
• pp.173-185
• /
• 2015

This study aims to investigate the characteristics and types of car accidents and establish a prediction model by analyzing 456 car accidents having occurred in the 11 tunnels in Busan, through statistical analysis techniques. The results of this study can be summarized as below. As a result of analyzing the characteristics of car accidents, it was found that 64.9% of all the car accidents took place in the tunnels between 08:00 and 18:00, which was higher than 45.8 to 46.1% of the car accidents in common roads. As a result of analyzing the types of car accidents, the car-to-car accident type was the majority, and the sole-car accident type in the tunnels was relatively high, compared to that in common roads. Besides, people at the age between 21 and 40 were most involved in car accidents, and in the vehicle type of the first party to car accidents, trucks showed a high proportion, and in the cloud cover, rainy days or cloudy days showed a high proportion unlike clear days. As a result of analyzing the principal components of car accident influence factors, it was found that the first principal components were road, tunnel structure and traffic flow-related factors, the second principal components lighting facility and road structure-related factors, the third principal factors stand-by and lighting facility-related factors, the fourth principal components human and time series-related factors, the fifth principal components human-related factors, the sixth principal components vehicle and traffic flow-related factors, and the seventh principal components meteorological factors. As a result of classifying car accident spots, there were 5 optimized groups classified, and as a result of analyzing each group based on Quantification Theory Type I, it was found that the first group showed low explanation power for the prediction model, while the fourth group showed a middle explanation power and the second, third and fifth groups showed high explanation power for the prediction model. Out of all the items(principal components) over 0.2(a weak correlation) in the partial correlation coefficient absolute value of the prediction model, this study analyzed variables including road environment variables. As a result, main examination items were summarized as proper traffic flow processing, cross-section composition(the width of a road), tunnel structure(the length of a tunnel), the lineal of a road, ventilation facilities and lighting facilities.

• Bulletin of the Korean Chemical Society
• /
• v.29 no.4
• /
• pp.833-841
• /
• 2008

Principal component-genetic algorithm-multiparameter linear regression (PC-GA-MLR) and principal component-genetic algorithm-artificial neural network (PC-GA-ANN) models were applied for prediction of melting point for 323 drug-like compounds. A large number of theoretical descriptors were calculated for each compound. The first 234 principal components (PC’s) were found to explain more than 99.9% of variances in the original data matrix. From the pool of these PC’s, the genetic algorithm was employed for selection of the best set of extracted PC’s for PC-MLR and PC-ANN models. The models were generated using fifteen PC’s as variables. For evaluation of the predictive power of the models, melting points of 64 compounds in the prediction set were calculated. Root-mean square errors (RMSE) for PC-GA-MLR and PC-GA-ANN models are 48.18 and $12.77{^{\circ}C}$, respectively. Comparison of the results obtained by the models reveals superiority of the PC-GA-ANN relative to the PC-GA-MLR and the recently proposed models (RMSE = $40.7{^{\circ}C}$). The improvements are due to the fact that the melting point of the compounds demonstrates non-linear correlations with the principal components.

• Journal of Pharmaceutical Investigation
• /
• v.37 no.6
• /
• pp.377-395
• /
• 2007

Near infrared(NIR) spectral data have been used for the noninvasive analysis of various biological samples. Nonetheless, absorption bands of NIR region are overlapped extensively. It is very difficult to select the proper wavelengths of spectral data, which give the best PCR(principal component regression) models for the analysis of constituents of biological samples. The NIR data were used after polynomial smoothing and differentiation of 1st order, using Savitzky-Golay filters. To find the best PCR models, all-possible combinations of available principal components from the given NIR spectral data were derived by in-house programs written in MATLAB codes. All of the extensively generated PCR models were compared in terms of SEC(standard error of calibration), $R^2$, SEP(standard error of prediction) and SECP(standard error of calibration and prediction) to find the best combination of principal components of the initial PCR models. The initial PCR models were found by SEC or Malinowski's indicator function and a priori selection of spectral points were examined in terms of correlation coefficients between NIR data at each wavelength and corresponding concentrations. For the test of the developed program, aqueous solutions of BSA(bovine serum albumin) and glucose were prepared and analyzed. As a result, the best PCR models were found using a priori selection of spectral points and the final model selection by SEP or SECP.

• IE interfaces
• /
• v.24 no.1
• /
• pp.8-14
• /
• 2011

Principal Component Analysis (PCA) reduces the dimensionality of the process by creating a new set of variables, Principal components (PCs), which attempt to reflect the true underlying process dimension. However, for highly nonlinear processes, this form of monitoring may not be efficient since the process dimensionality can't be represented by a small number of PCs. Examples include the process of semiconductors, pharmaceuticals and chemicals. Nonlinear correlated process variables can be reduced to a set of nonlinear principal components, through the application of Kernel Principal Component Analysis (KPCA). Support Vector Data Description (SVDD) which has roots in a supervised learning theory is a training algorithm based on structural risk minimization. Its control limit does not depend on the distribution, but adapts to the real data. So, in this paper proposes a non-linear process monitoring technique based on supervised learning methods and KPCA. Through simulated examples, it has been shown that the proposed monitoring chart is more effective than $T^2$ chart for nonlinear processes.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.24 no.5
• /
• pp.1013-1022
• /
• 2014

Correlation Power Analysis (CPA) is a type of Side-Channel Analysis (SCA) that extracts the secret key using the correlation coefficient both side-channel information leakage by cryptography device and intermediate value of algorithms. Attack performance of the CPA is affected by noise and temporal synchronization of power consumption leaked. In the recent years, various researches about the signal processing have been presented to improve the performance of power analysis. Among these signal processing techniques, compression techniques of the signal based on Principal Component Analysis (PCA) has been presented. Selection of the principal components is an important issue in signal compression based on PCA. Because selection of the principal component will affect the performance of the analysis. In this paper, we present a method of selecting the principal component by using the correlation of the principal components and the power consumption is high and a CPA technique based on the principal component that utilizes the feature that the principal component has different. Also, we prove the performance of our method by carrying out the experiment.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.10a
• /
• pp.69-72
• /
• 2003

The paper considers partial least squares (PLS) as a new dimension reduction technique for the feature vector to overcome the small sample size problem in face recognition. Principal component analysis (PCA), a conventional dimension reduction method, selects the components with maximum variability, irrespective of the class information. So PCA does not necessarily extract features that are important for the discrimination of classes. PLS, on the other hand, constructs the components so that the correlation between the class variable and themselves is maximized. Therefore PLS components are more predictive than PCA components in classification. The experimental results on Manchester and ORL databases show that PLS is to be preferred over PCA when classification is the goal and dimension reduction is needed.

• Korean Institute of Interior Design Journal
• /
• v.21 no.4
• /
• pp.163-171
• /
• 2012

The objective of this study was to analyze the characteristics of traditionality expressions at modernized Chinese restaurant in Hong Kong. As a case study, the study examined 12 modernized Chinese restaurants in Hong Kong. The gathered data were categorized and examined according to the ways of traditionality expressions, which included reproduction, transformation, and reinterpretation of traditional components. Each of the components was measured for the amount of traditional or modernity expression on a five-point scale. The five-point scoring system put an emphasis on heritage; 1 point was given to principal modernity(modernity: 90-100% + tradition: 0-10%), 2 points were given to principal modernity + auxiliary tradition(modernity: 70-90% + tradition: 10-30%), 3 points were given to the same ratio between tradition and modernity(modernity: 40-60% + tradition: 40-60%), 4 points were given to principal tradition + auxiliary modernity(modernity: 10-30% + tradition: 70-90%), and 5 points were given to principal tradition(modernity: 0-10% + tradition: 90-100%). The analysis performed according to those criteria and methodologies led to the following findings and conclusions: Traditional components were most reproduced in the ornaments placed all over the restaurant and applied to the chirography of the restaurant logos, walls, and windows/doors in a big number. The methodology of transforming tradition was evenly applied to each of the spatial components. With the most transformations occurring to the lattices, there were many different ways to transform tradition including the partition, chirography, pattern, red lantern, furniture and ornament, and traditional materials that were turned into modern ones. Few examples of reinterpreting tradition were observed in the restaurant titles, inside floors, and ceilings, but plenty of examples were found in the walls, windows/doors, lighting, and furniture in a range of ways. Most of them reinterpreted the traditional forms and added altered patterns to them to remind customers of tradition. In short, all of the three ways of expressing tradition were actively applied to each component in an array of ways.