• 한국해양공학회지
• /
• 제26권4호
• /
• pp.57-63
• /
• 2012

In this paper, a Bayesian classifier based on PCA (principle component analysis) is proposed to classify underwater transient signals using $16^{th}$ order LPC (linear predictive coding) coefficients as feature vector. The proposed classifier is composed of two steps. The mechanical signals were separated from biological signals in the first step, and then each type of the mechanical signal was recognized in the second step. Three biological transient signals and two mechanical signals were used to conduct experiments. The classification ratios for the feature vectors of biological signals and mechanical signals were 94.75% and 97.23%, respectively, when all 16 order LPC vector were used. In order to determine the effect of underwater noise on the classification performance, underwater ambient noise was added to the test signals and the classification ratio according to SNR (signal-to-noise ratio) was compared by changing dimension of feature vector using PCA. The classification ratios of the biological and mechanical signals under ocean ambient noise at 10dB SNR, were 0.51% and 100% respectively. However, the ratios were changed to 53.07% and 83.14% when the dimension of feature vector was converted to three by applying PCA. For correct, classification, it is required SNR over 10 dB for three dimension feature vector and over 30dB SNR for seven dimension feature vector under ocean ambient noise environment.

• 전기학회논문지
• /
• 제65권9호
• /
• pp.1557-1563
• /
• 2016

According to recent studies, poor sitting posture of the spine has been shown to lead to a variety of spinal disorders. For this reason, it is important to measure the sitting posture. We proposed a strategy for classification of sitting posture using machine learning. We retrieved acceleration data from single tri-axial accelerometer attached on the back of the subject's neck in 5-types of sitting posture. 6 subjects without any spinal disorder were participated in this experiment. Acceleration data were transformed to the feature vectors of principle component analysis. Support vector machine (SVM) and K-means clustering were used to classify sitting posture with the transformed feature vectors. To evaluate performance, we calculated the correct rate for each classification strategy. Although the correct rate of SVM in sitting back arch was lower than that of K-means clustering by 2.0%, SVM's correct rate was higher by 1.3%, 5.2%, 16.6%, 7.1% in a normal posture, sitting front arch, sitting cross-legged, sitting leaning right, respectively. In conclusion, the overall correction rates were 94.5% and 88.84% in SVM and K-means clustering respectively, which means that SVM have more advantage than K-means method for classification of sitting posture.

• 농촌계획
• /
• 제21권2호
• /
• pp.103-114
• /
• 2015

The comprehensive rural village development projects (CRVDP) have been carried out as the core one of the rural development schemes in Korea since 2004. CRVDP included the various rural experience programs to increase rural income and in order to promote rural community development in the project area. This study analyzed the operating management conditions, types and characteristics of the rural experience programs targeting the 168 CRVDPs have been completed so that the recommendations and lessons which were found the usefulness, challenges and improvements to the CRVDP can be provided to be better the same kinds of rural development projects. We identified the relationships between performances such as increasing village income and utilization of rural amenity resources to the CRVDP and operational management types of the rural experience programs as well. Employing principle component analysis and cluster analysis technique, this study found 5 clusters of rural experience programs among 168 CRVDPs. The results of analysis of variance indicated that there were significant the mean differences between clusters such as the utilization of rural amenity resources(0.01), income of rural experience programs(0.1). According to the result of the Chi-squire test, there was very significant differences between internet homepage operation and clusters(0.01). Finally, the analysis of covariance about the income of rural experience programs showed that there were significant the mean differences between clusters(0.05).

• 복식
• /
• 제45권
• /
• pp.71-84
• /
• 1999

This study was to identify the dimensions of consumer's purchasing motives and purchasing delay reasons toward casual hanbok. Other objective was to examine relationship between these variables and future purchasing intention. Th 22 purchasing motive questions and 19 purchasing delay reasons were selected through the result of self-questionnaire analysis. 302 purchaser and 297 consumers who delay for particular reasons in Pusan responsed to the second questionnaire of purchasing motives and purchasing delay reasons toward casual Hanbok. The results as follows: 1. For factor analysis 22 purchasing motive questions were subjected to the principle component analysis with orthogonal rotation after extraction of 6 major factors. Six dimensions are consciousness of nation goodness of design conformity with fashion charming apperance relaxation fo body and mind nation goodness of design conformity with fashion charming appearance relaxation of body and mind and pursuit of individuality. These factors explained 62.0% of total variables. 2. Consumer's purchasing motives such as consciousness of nation goodness of design charming appearance and relaxation of body and mind has predicting power to the re-purchasing intention of casual hanbok 3. For factor analysis 19 delay reason question were subjected to the principle component analysis with orthogonal rotation after extraction of 5 major factors. Five dimensions are non-fitness for occasion and body shape unsatisfaction with design unsatisfaction with price need of information search for better product and preference for traditional hangok. These factors explained 60.4% of total variables. 4. Delay reasons of unsatisfaction with design and price were positively related to the future purchasing intention. This delay reason is caused by forces external to the consumer and the consumer has engaged in information search. This result explained this type of consumer has the strong future purchasing intention.

• 전자공학회논문지B
• /
• 제33B권1호
• /
• pp.96-107
• /
• 1996

본 논문에서는 직교 벡터 공간 변환을 이용한 새로운 음성 개성 변환 알고리즘을 제안하였다. 음성 개성 변환이란 임의 환자(source)가 가지고 있는 몇 개의 특징 변수를 다른 화자(target)의 특징 변수로 변환하는 기법이다. 본 논문에서는 LPC 켑스트럼 계수와 여기 신호의 스펙트럼, 그리고 피치 궤적을 변환하여 음성 개성변환을 구현하였다. LPC 켑스트럼 계수의 변환을 위해 직교 벡터 공간 변환 기법이 제안되었다. 이 기법은 KL(Karhunen-Loeve)변환을 이용한 principle component의 분리와 최소 자승 오차를 갖는 선형 좌표 변환을 통해 LPC 켑스트럼의 변환을 수행한다. 또한, 화자간의 운율적인 특징을 변환하기 위해 피치 궤적 변환 기법이 제안되었다. 피치 궤적 변환을 위하여 먼저 두 화자간의 기준 피치 패턴의 작성하고 기준 패턴간의 대응 관계를 추정한 후 이를 이용하여 source 화자의 피치 패턴이 target 피치 패턴으로 변환되도록 하였다. 컴퓨터를 이용한 모의 실험 결과 제안된 알고리즘은 객관적인 평가와 주관적인 평가에 있어서 우수한 성능을 나타내었다.

• Asian-Australasian Journal of Animal Sciences
• /
• 제27권3호
• /
• pp.303-309
• /
• 2014

The increasing importance of meat quality has implications for animal breeding programs. Research has revealed much about the genetic background of pigs, and many studies have revealed the importance of various genetic factors. Since meat quality is a complex trait which is affected by many factors, consideration of the overall phenotype is very useful to study meat quality. For integrating the phenotypes, we used principle component analysis (PCA). The significant SNPs refer to results of the GRAMMAR method against PC1, PC2 and PC3 of 14 meat quality traits of 181 Duroc pigs. The Genome-wide association study (GWAS) found 26 potential SNPs affecting various meat quality traits. The loci identified are located in or near 23 genes. The SNPs associated with meat quality are in or near five genes (ANK1, BMP6, SHH, PIP4K2A, and FOXN2) and have been reported previously. Twenty-five of the significant SNPs also located in meat quality-related QTL regions, these result supported the QTL effect indirectly. Each single gene typically affects multiple traits. Therefore, it is a useful approach to use integrated traits for the various traits at the same time. This innovative approach using integrated traits could be applied on other GWAS of complex-traits including meat-quality, and the results will contribute to improving meat-quality of pork.

• Journal of Community Nutrition
• /
• 제6권3호
• /
• pp.121-124
• /
• 2004

Objective : The paper aims to provide predictive coefficients via BIA for the assessment of body composition in children and adolescents to serve clinical as well as research purposes. Methods : Body composition via dual-energy x­ray absorptiometry (DXA) and bioelectric impedance as well as other anthropometric index were derived from meaurements on 1026 children and adolescents aged from 6 to 18 years from Beijing City. The best subset regression and principle component analysis were adopted to build the predictive coefficients with the logarithm of body composition via DXA as response variable. Results : Condition index ${\varphi}$ of fat-free mass multiple linear regression achieves 113.49 and 91.18 for males and females respectively, demonstrating severe multicollinearity among anthropometric indexes in children and adolescents. BIA predictive coefficients base on the best subset regression and principle component analysis boast a content predictive value for lean mass ($r^2$ = 0.9697 and 0.9664 for boys and girls respectively, p < 0.0001) and for Fat$\%$ ($r^2$ = 0.7705 and 0.6959 for boys and girls respectively, p < 0.0001). Conclusions : BIA method is applicable for the prediction of body composition for children and adolescents.

• 인터넷정보학회논문지
• /
• 제10권5호
• /
• pp.183-194
• /
• 2009

본 논문에서는 인터넷상에 업로드되는 음란 영상물을 차단하기 위해 활동적 형태 모델(active shape model)을 이용한 유해 영상 탐지 방법을 제안한다. 본 논문에서는 활동적 형태 모델을 이용하여 가슴선의 형태를 주성분 분석(Principle Component Analysis)과 정렬을 통해서 학습하고, 각 제어점에 대응하는 화소값 분포를 학습한다. 그리고 학습된 형태와 화소값 분포를 이용하여 가슴선을 찾는다. 본 논문에서는 형태 모델의 초기 위치를 정확하게 선택하기 위해 스케일, 회전, 이동에 관한 파라미터를 추출한다. 이 정보를 획득하기 위해서 본 논문에서는 유두 부분의 위치를 찾고, 유두 위치로부터 모든 방향으로 방사하여 후보 가슴선을 찾는다. 이와 같이 검출한 가슴선 정보를 이용하여 스케일과 회전 값을 찾아 평균 형태(mean shape)를 위치시키고, 활동적 형태 모델을 반복적으로 탐색한다. 최종적으로 수렴한 형태의 제어점(landmark)과 후보 가슴선과의 거리 평균을 계산하여 유해영상의 유무를 판단한다.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제32권12호
• /
• pp.1247-1259
• /
• 2005

본 연구는 웹카메라와 같은 저해상도의 동영상으로부터 실시간 다중 얼굴 인식 시스템을 제안한다. 동영상을 이용한 얼굴 인식 시스템은 크게 얼굴 검출 단계와 얼굴 분류 단계로 나눌 수 있다. 첫째, 얼굴 검출 단계에서는 빠르고 강인한 객체 검출 성능을 가진 AdaBoost를 이용하여 얼굴 후보 영역을 검출하였고, 검출된 얼굴 후보 영역에 대한 주성분을 수행하여 데이타의 크기기 현저히 줄어든 특징 벡터를 구한 다음에 특징 벡터에 대해 SVM 기반 이진 분류를 수행하여 얼굴 후보 영역을 검증하였다. 둘째, 얼굴 분류 단계에는 주성분 분석과 멀티 SVM을 이용하여 각 얼굴들을 분류하였다. 실험 결과 본 논문에서 제안한 방법은 저해상도에서도 높은 얼굴 검출율과 동영상에서 실시간 처리가 가능한 빠른 다중 얼굴 검출과 인식 성능을 보였다. 또한 팬-틸트 기능을 가진 웹카메라를 이용한 자동 추적형 얼굴 인식 시스템을 적용하여 얼굴 검출 성능을 향상시켰고, 얼굴 인식 시스템의 응용으로 무선 On/off 얼굴인식 도어락 시스템을 구현하였다.

• 대한원격탐사학회지
• /
• 제35권6_3호
• /
• pp.1173-1185
• /
• 2019

본 연구에서는 처음으로 주성분분석(Principle component analysis; PCA) 방법을 이용하여 Sentinel-5p의 TROPOspheric Monitoring Instrument (TROPOMI) 위성센서 원시자료로부터 산업활동 및 화산활동에 의해 발생한 이산화황 연직칼럼농도(Vertical column density; VCD)를 산출하였다. 본 연구에서 TROPOMI로부터 주성분분석방법을 이용하여 산출된 이산화황 연직칼럼농도는 차등흡수분광법(Differential Optical Absorption Spectroscopy; DOAS)을 이용하여 산출된 TROPOMI Level 2 이산화황 연직칼럼농도 산출물과 비교되었다. 산업활동과 같은 인위적 요인에 의하여 다량의 이산화황을 지표부근에 배출하는 동아시아 지역에서 TROPOMI 로부터 주성분분석방법으로 산출된 이산화황 연직칼럼농도와 차등흡수분광법을 이용하여 산출된 TROPOMI 이산화황 연직칼럼농도의 평균값은 각각 0.05 Dobson Unit (DU)와 -0.02 DU로 비슷한 값으로 나타났다. 두 산출물 사이의 기울기(Slope)는 모든 구름조건에 대하여 0.64, 상관계수(Correlation coefficient, R)는 0.51로 다소 낮은 상관관계를 보였으나, 구름비율이 0.5 이하인 픽셀에 대한 기울기는 0.68, 상관계수는 0.61로 증가하였다. 이러한 결과는 두 알고리즘에서 공통적으로 구름비율이 높을 때 지표부근에 대한 이산화황의 산출 민감도가 감소한다는 것을 의미한다. 화산활동에 의한 고농도 이산화황이 발생하는 지역인 인도네시아와 일본 남부 지역에서 두 알고리즘으로 산출된 이산화황 연직칼럼농도 사이의 상관계수는 모든 구름 조건에 대하여 0.90으로 높은 상관관계를 보였다. 이는 화산지역에서의 가스 분출로 인하여 고농도로 대류권 상층 혹은 성층권 하부에 주로 분포하는 이산화황에 대한 위성 기반 이산화황 산출 정확도가 높게 나타나기 때문인 것으로 사료된다.