• 대한가정학회지
• /
• 제42권11호
• /
• pp.85-107
• /
• 2004

This study classified figure types of adult males into several kinds of shape to provide fundamental data for their clothing sizing system. The subjects were 1496 men aged between 20 and 60 years old. Data were analyzed by factor analysis, cluster analysis and discriminant analysis. The results were as follows 1. For the result of the interview, the data were grouped into three age brackets: 20-35,31-45 and 41-60 years. 2. Factor analysis using values, which were measurements divided by either weight or height, was carried out to extract factors which characterize the various figures. fve factors to determine the figure types were extracted. 3. Cluster analysis using factor scores was carried out to categorize the figure types within the age groups. Figure types, describing shoulder angie and body shape, were categorized into 3 per age group. 4. Stepwise discriminant analysis w3s used to ensure that these clusters could be utilized with appropriate hit ratio. The hit ratio for each age group was around 80%.

• 한국의류학회지
• /
• 제26권1호
• /
• pp.15-26
• /
• 2002

The purpose of this study was to classier the somatotype of late middle-aged women and to analyze the characteristics of each somatotype. The subjects were 337 late middle-aged women and their age range os from 45 to 59 fears old. Data were collected through anthropometry and photometry and analyzed by factor analysis, cluster analysis and discriminant analysis. The results were as follows; 1. The result of factor analysis indicated that 9 factors were extracted through factor analysis and those factors comprised 83.56 percent of total valiance. 2. Using factor scores, cluster analysis was carried out and the subject were classified into 4 cluster. Each cluster was classified as their body front and side view contour. Type 1 is tall, slim, and lower balk is flat on the side. Type 2 is standard and lean-back type on the side. Type 3 is standard height and weight, H type in front, and belly-protruded on the side. Type 4 is short, fat, and the side is hip-protruded. 3. According to the stepwise discriminant analysis, the 9 important items in classifying the somatotype of the late middle-aged women are as follows ; lower back tilt angle, hip depth(back) -back waist depth(back), bust depth(fore) - anterior waist depth(fore), jugular fossa point(fore), upper back tilt angle, burst breadth -waist breadth, right shoulder tilt, height of shoulder - height of anterior waist, abdomen breath. The correct classification rate for these items is as exact as 84.62%.

• 한국의류학회지
• /
• 제24권3호
• /
• pp.345-352
• /
• 2000

The purpose of this study was to classify lower body types of elementary school girls. The subjects for anthropometric study were 368 girls aged from 10 to 11. Factor analysis, cluster analysis, discriminant analysis, and analysis of variance were performed for statistical analysis of the data. Four lower body construction factors were extracted by the factor analysis of antropometric measurements. The factors extracted were lower body fatness factor, lower body height factor, lower body length from the waist to the crotch factor, and lower body configuration factor. On the basis of the cluster analysis, three different lower body types were categorized. Type 1 was short and small sized type and 42.4% of subjects belonged under this type. Type 2 was tall and fat type and 22.3% of subjects belonged under this type. Type 3 was the most similar to the average type having the largest waist-hip drop value and 35.3% of subjects belonged under this type. Discriminant analysis showed 7 discriminant factors that can classify the children's lower body type were Rohrer's index, height, fibulae length, waist girth, ilio cristale girth, trochanter girth, and weight.

• 품질경영학회지
• /
• 제29권3호
• /
• pp.124-137
• /
• 2001

In general, it is very efficient and effective to use screening variables that are correlated with the performance variable in case that measuring the performance variable is impossible (destructive) or expensive. The general methodology for searching surrogate variables is regression analysis. This paper considers the inspection problem in CRT (Cathode Ray Tube) production line, in which the performance variable (dependent variable) is binary type and screening variables are continuous. The general regression with dummy variable, discriminant analysis and binary logistic regression are considered. The cost model is also formulated to determine economically inspection procedure with screening variables.

• Bulletin of the Korean Chemical Society
• /
• 제7권5호
• /
• pp.347-353
• /
• 1986

Fisher's discriminant method has been applied to the problem of the classification of Korean potsherds, using their elemental composition as analyzed by neutron activation analysis. A combination of analytical data by means of statistical linear discriminant analysis has resulted in removal of redundant variables, optimal linear combination of meaningful variables and formulation of classification rules.

• Journal of the Korean Statistical Society
• /
• 제30권2호
• /
• pp.247-263
• /
• 2001

This paper concerns a problem of classification based on training dta. A framework of information geometry is given to elucidate the characteristics of discriminant functions including logistic discrimination and AdaBoost. We discuss a class of loss functions from a unified viewpoint.

• Journal of the Korean Statistical Society
• /
• 제21권1호
• /
• pp.1-13
• /
• 1992

This article deals with variable selection problem under a newly formed predictive heteroscedastic discriminant rule that accounts for mulitple homogeneous covariance matrices across the K multivariate normal populations. A general version of predictive discriminant rule, a variable selection criterion, and a criterion for stopping with further selection are suggested. In a simulation study the practical utilities of those considered are demonstrated.

• Journal of Biosystems Engineering
• /
• 제24권4호
• /
• pp.325-334
• /
• 1999

The aim of this study was to present a groundwork for development of a sorting system of peeled garlics using machine vision. Images of various garlic samples such as sound, partially defective, discolored, rotten and un-peeled were obtained with a B/W machine vision system. Sorting factors which were based on normalized histogram and statistical analysis(STEPDISC Method) had good separability for various garlic samples. Bayes discriminant function and neural network sorting algorithms were developed with the sample images and were experimented on various garlic samples. It was showed that garlic samples could be classified by sorting algorithm with average sorting accuracies of 88.4% by Bayes discriminant function and 93.2% by neural network.

• 한국산학기술학회논문지
• /
• 제21권2호
• /
• pp.359-365
• /
• 2020

한방병원에 건강검진을 목적으로 내원한 732명의 피험자를 대상으로 사상체질감별설문지 검사를 통하여 한의사가 감별한 사상체질 결과와 맥진기를 사용하여 촌, 관, 척 부위의 요골 맥파를 측정한 자료를 입수하였다. 입수한 맥파 자료로부터 맥파의 폭, 피크의 크기와 개수를 특징 변수로 추출했다. 특징 변수들 중 타당도와 신뢰도가 높은 것들을 판별변수로 선정했다. 5겹 교차 검증법을 적용하여, 맥파 자료를 사상체질을 알려주는 훈련 표본과 사상체질을 알려주지 않는 예측 표본으로 구분했다. 훈련 표본에 대해서 판별분석을 수행하여 판별 함수를 구한 후, 이것을 예측 표본에 적용하여 사상체질을 예측했다. 예측한 사상체질과 설문지 검사로 알고 있는 사상체질을 비교하여 사상체질 분류 정확도를 계산했다. 사상체질 분류 정확도는 나이별, 성별 보정 전에는 태음인: 73.6 %, 소음인: 68.4 %, 소양인: 74.2 %이었으며, 전체 분류 정확도는 72.5 %이었다. 한국 인구를 기반으로 한 성별, 나이별 가중치를 적용하여 보정을 한 후의 사상체질 분류 정확도는 태음인(70.4 %), 소음인(84.2 %), 소양인(67.7 %)이었으며, 전체 분류 정확도는 73.8 %이었다.

• 한국통신학회논문지
• /
• 제34권10B호
• /
• pp.1111-1116
• /
• 2009

본 논문에서는 집적 영상의 획득과 복원을 이용하여 왜곡에 강인한 물체를 인식하는 방법을 연구한다. 해당 화소들의 확률적 특성인 평균과 표준편차를 이용하여 3차원 공간에서 물체를 복원하고 거리를 추정한다. 표적인식은 Fisher 선형판별법(linear discriminant analysis, LDA)과 주성분 분석법(principal component analysis, PCA) 기술을 결합한 통계적 분류기(statistical classifier)로 수행한다. Fisher 선형판별법은 클래스 간의 판별력을 최대로 하고 주성분 분석법은 Fisher 선형판별법을 수행하기 위한 차원축소를 실행한다. 주성분 분석법은 차원축소 후 복원된 벡터와 원 벡터의 오차를 최소화하는 기술로 알려져 있다. 실험 및 시뮬레이션을 통하여 면외(out-of-plane) 회전된 표적을 본 논문에서 제안한 방법으로 분류한다.