• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.1C
• /
• pp.65-71
• /
• 2004

We improved the MLLR speaker adaptation algorithm with reduction of the order of HMM parameters using PCA(Principle Component Analysis) or ICA(Independent Component Analysis). To find a smaller set of variables with less redundancy, we adapt PCA(principal component analysis) and ICA(independent component analysis) that would give as good a representation as possible, minimize the correlations between data elements, and remove the axis with less covariance or higher-order statistical independencies. Ordinary MLLR algorithm needs more than 30 seconds adaptation data to represent higher word recognition rate of SD(Speaker Dependent) models than of SI(Speaker Independent) models, whereas proposed algorithm needs just more than 10 seconds adaptation data. 10 components for ICA and PCA represent similar performance with 36 components for ordinary MLLR framework. So, compared with ordinary MLLR algorithm, the amount of total computation requested in speaker adaptation is reduced by about 1/167 in proposed MLLR algorithm.

• KOREAN JOURNAL OF CROP SCIENCE
• /
• v.43 no.2
• /
• pp.101-104
• /
• 1998

This study was conducted to establish a rapid analysis method for determining protein and moisture contents of pea. Ninety and eighty pea (Pisum sativum L.) lines were analyzed to determine protein and moisture contents, respectively using near-infrared reflectance spectroscopy. Simple correlations (${\gamma}$) of protein content in a ground sample and an intact grain sample by an automatic regression method were 0.978 and 0.910, respectively. Simple correlations by partial least square regression/principal component analysis (PLS/PCA) methods were 0.982 and 0.925, respectively. Standard error of performance (SEP) in protein content was the lowest value, 0.446 in ground sample by PLS/PCA methods. Simple correlation of moisture content was the highest at 0.871 in ground samples. when using a standard regression method. Accuracy for the moisture content was slightly lower than for protein content. It was concluded that the NIRS method would be applicable only for rapid determination of protein content in pea.

• Proceedings of the Korean Society of Near Infrared Spectroscopy Conference
• /
• 2001.06a
• /
• pp.1241-1241
• /
• 2001

We propose a new calibration method, which uses the linearization method for spectral responses and the repetitive adoptions of the linearization weight matrices to construct a frature. Weight matrices are estimated through multiple linear regression (or principal component regression or partial least squares) with forward variable selection. The proposed method is applied to three data sets. The first is FTIR spectral data set for FeO content from sinter process and the second is NIR spectra from trans-alkylation process having two constituent variables. The third is NIR spectra of crude oil with three physical property variables. To see the calibration performance, we compare the new method with the PLS. It is found that the new method gives a little better performance than the PLS and the calibration result is stable in spite of the collinearity among each selected spectral responses. Furthermore, doing the repetitive adoptions of linearization matrices in the proposed methods, uninformative variables are disregarded. That is, the new methods include the effect of variables subset selection, simultaneously.

• Journal of Electrical Engineering and Technology
• /
• v.12 no.4
• /
• pp.1406-1416
• /
• 2017

In complex and large-scale industries, properly designed fault detection and identification (FDI) systems considerably improve safety, reliability and availability of target processes. In thermal power plants (TPPs), generating units operate under very dangerous conditions; system failures can cause severe loss of life and property. In this paper, we propose a bagged auto-associative kernel regression (AAKR)-based FDI approach for steam boilers in TPPs. AAKR estimates new query vectors by online local modeling, and is suitable for TPPs operating under various load levels. By combining the bagging method, more stable and reliable estimations can be achieved, since the effects of random fluctuations decrease because of ensemble averaging. To validate performance, the proposed method and comparison methods (i.e., a clustering-based method and principal component analysis) are applied to failure data due to water wall tube leakage gathered from a 250 MW coal-fired TPP. Experimental results show that the proposed method fulfills reasonable false alarm rates and, at the same time, achieves better fault detection performance than the comparison methods. After performing fault detection, contribution analysis is carried out to identify fault variables; this helps operators to confirm the types of faults and efficiently take preventive actions.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.2
• /
• pp.13-18
• /
• 2010

Today, The much amount of spam that occupies the mail server and network storage occurs the lack of negative issues, such as overload, and for users to delete the spam should spend time, resources have a problem. Automatic spam filtering on the incidence to solve the problem is essential. A lot of Spam filters have tried to solve the problem emerged as an essential element automatically. Unlike traditional method such as Naive Bayesian, PCA through the many-dimensional data set of spam with a few spindle-dimensional process that narrowed the operation to reduce the burden on certain groups for classification Logistic regression analysis method was used to filter the spam. Through the speed and performance, it was able to get the positive results.

• Korean Journal of Soil Science and Fertilizer
• /
• v.43 no.5
• /
• pp.748-754
• /
• 2010

The contents of soil organic matter (SOM) and $Al_o+1/2Fe_o$ in soils are important criteria for the classification of new Andisols in Soil Taxonomy system. There are many soil types in Jeju Island with various soil forming environments. This paper was conducted to estimate the contents of soil organic matter and the content of ammonium oxalate extracted Al and Fe ($Al_o+1/2Fe_o$) using various environmental variables and to make soil property maps using a statistical analyses. The soil samples were collected from 321 locations and analyzed to measure the contents of SOM and $Al_o+1/2Fe_o$. It was analyzed the relationships among them and various environmental variables such as temperature, precipitation, net primary product, radiation, evapotranspiration, altitude, soil forming energy, topographic wetness index, elevation, difference surrounded area, and distances from the shore and the peak. We can exclude multi-collinearity among environmental variables with principal component analysis and reduce all the variables to 3 principal components. The contents of SOM and $Al_o+1/2Fe_o$ were estimated by multiple regression models and maps of them were made using the models.

• Journal of Ecology and Environment
• /
• v.33 no.2
• /
• pp.165-173
• /
• 2010

Soil respiration ($R_S$) is a critical component of the annual carbon balance of forests, but few studies thus far have attempted to evaluate empirical regression models in $R_S$. The principal objectives of this study were to evaluate the relationship between $R_S$ rates and soil temperature (ST) and soil water content (SWC) in soil from a cool-temperate deciduous broad-leaved forest, and to evaluate empirical regression models for the prediction of $R_S$ using ST and SWC. We have been measuring $R_S$, using an open-flow gas-exchange system with an infrared gas analyzer during the snowfree season from 1999 to 2001 at the Takayama Forest, Japan. To evaluate the empirical regression models used for the prediction of $R_S$, we compared a simple exponential regression (flux = $ae^{bt}$Eq. [1]) and two polynomial multiple-regression models (flux = $ae^{bt}{\times}({\theta}{\nu}-c){\times}(d-{\theta}{\nu})^f:$ Eq. [2] and flux = $ae^{bt}{\times}(1-(1-({\theta}{\nu}/c))^2)$: Eq. [3]) that included two variables (ST: t and SWC: ${\theta}{\nu}$) and that utilized hourly data for $R_S$. In general, daily mean $R_S$ rates were positively well-correlated with ST, but no significant correlations were observed with any significant frequency between the ST and $R_S$ rates on periods of a day based on the hourly $R_S$ data. Eq. (2) has many more site-specific parameters than Eq. (3) and resulted in some significant underestimation. The empirical regression, Eq. (3) was best explained by temporal variations, as it provided a more unbiased fit to the data compared to Eq. (2). The Eq. (3) (ST $\times$ SWC function) also increased the predictive ability as compared to Eq. (1) (only ST exponential function), increasing the $R^2$ from 0.71 to 0.78.

• Journal of the Korea Convergence Society
• /
• v.10 no.12
• /
• pp.271-280
• /
• 2019

Smart factory, a critical part of digital transformation, enables data-driven decision making using monitoring, analysis and prediction. Predictive maintenance is a key element of smart factory and the need is increasing. The purpose of this study is to analyze the degradation characteristics of a galvanizing kettle for the steel plating process and to predict the remaining useful life(RUL) for predictive maintenance. Correlation analysis, multiple regression, principal component regression were used for analyzing factors of the process. To identify the trend of degradation, a proposed rolling window was used. It was observed the degradation trend was dependent on environmental temperature as well as production factors. It is expected that the proposed method in this study will be an example to identify the trend of degradation of the facility and enable more consistent predictive maintenance.

• Journal of the Korean Society of Food Culture
• /
• v.35 no.2
• /
• pp.193-200
• /
• 2020

This study was conducted to understand the sensory characteristics and consumer acceptance for the commercially available clear broth for noodles. Totally, eight different clear broth samples were evaluated in this study. Seven trained panelists developed and evaluated sensory characteristics in the descriptive analysis. Significant differences (p<0.05) were obtained for all 28 attributes evaluated. Descriptive data was obtained by performing multivariate analysis of variance to identify differences between samples. Principal component analysis (PCA) was performed on the mean values of descriptive attributes obtained in the descriptive analysis, and summarizes the sensory characteristics of clear broth for noodles. PCA of the clear broths revealed that the first two principal components are responsible for 80.66% variations. For sensory testing, 160 consumers were recruited, and their acceptance for each sample was assessed. Consumer data was obtained by applying partial least square-regression (PLSR) to establish the relationship between the descriptive data and the consumer acceptance data.

• Journal of Korean Academy of Nursing Administration
• /
• v.15 no.1
• /
• pp.64-71
• /
• 2009

Purpose: This study shows the prehospital emergency nursing practices, and analyzes them associated with their individual characteristics and job conditions. Method: Based on a survey of the National Emergency Medical Center in Korea(2008), principal components were extracted from 7 prehospital emergency nursing practices by factor analysis, and some regression analyses of principal components(CPR-AED and V/S-I.V.) were executed on individual characteristics and job conditions. Results: The PENs gave themselves higher order ratings for vital sign check, airway management for loss of consciousness patients, CPR for suspicious cardiac arrest, keeping vein open for shock patients, AED for abnormal pulse rate, AED for suspicious cardiac arrest, and AED for loss of consciousness. Age and duty periods were statistically significant influential factors on the CPR-AED component. Conclusion: The results indicate that the PENs were competent in overall prehospital emergency activities and procedures even some weak self-evaluations, and that the standard curriculum and practice standard for prehospital nursing should be developed in order to increase nursing leadership in prehospital emergency settings.