• 제목/요약/키워드: canonical correlation analysis

검색결과 153건 처리시간 0.115초

Higher-order solutions for generalized canonical correlation analysis

  • Kang, Hyuncheol
    • Communications for Statistical Applications and Methods
    • /
    • 제26권3호
    • /
    • pp.305-313
    • /
    • 2019
  • Generalized canonical correlation analysis (GCCA) extends the canonical correlation analysis (CCA) to the case of more than two sets of variables and there have been many studies on how two-set canonical solutions can be generalized. In this paper, we derive certain stationary equations which can lead the higher-order solutions of several GCCA methods and suggest a type of iterative procedure to obtain the canonical coefficients. In addition, with some numerical examples we present the methods for graphical display, which are useful to interpret the GCCA results obtained.

Nonlinear Canonical Correlation Analysis for Paralysis Disease Data

  • Shin, Yang-Kyu
    • Journal of the Korean Data and Information Science Society
    • /
    • 제15권3호
    • /
    • pp.515-521
    • /
    • 2004
  • Categorical data are mostly found in oriental medical research. The nonlinear canonical correlation analysis does not assume an interval level of measurement. In this paper, we apply nonlinear canonical correlation analysis to quantification and explain how similar sets of variables are to one another for paralysis disease data.

  • PDF

정준상관분석을 통한 다변량 금융시계열의 변동성 분석 (Multivariate Volatility Analysis via Canonical Correlations for Financial Time Series)

  • 이승연;황선영
    • 응용통계연구
    • /
    • 제27권7호
    • /
    • pp.1139-1149
    • /
    • 2014
  • 다변량 금융시계열의 변동성분석을 다변량 기법인 정준상관분석(canonocal correaltion analysis)을 이용해 분석하였다. 변동성의 특성상 계수들이 비음(non-negative)인 정준상관분석, 즉, non-negative and sparse canonical correlation analysis (NSCCA)를 이용해 보았다. 본 논문은 다변량 시계열의 변동성 커브에 대해 연구하고 있으며 제시된 방법론을 이변량 주식자료분석을 통해 예시해 보았다.

정준상관분석을 이용한 수중표적 분석 (Underwater Target Analysis Using Canonical Correlation Analysis)

  • 석종원;김태환;배건성
    • 한국정보통신학회논문지
    • /
    • 제16권9호
    • /
    • pp.1878-1883
    • /
    • 2012
  • 일반적으로 수중표적 인식에서는 표적의 형상/재질에 따른 수신 표적신호의 공간적인 정보를 특징인자로 추출하여 식별하고자 하는 특징을 추출하였다. 또한, 표적신호의 수신 위치에 덜 민감한 특징파라미터 추출을 위해 다양한 신호처리 기법을 적용하는 연구가 수행되어 왔다. 본 논문에서는 표적신호의 수신위치에 상대적으로 민감하지 않은 정준상관분석(Canonical correlation Analysis; CCA)을 사용하여 합성된 수중물체의 특징을 분석하였다. 다중각도 환경에서 특징추출을 위해 정준산관분석기법이 적용되었으며, 각각 다른 각도에서 수중물체에 반사되어 되돌아오는 연속적인 두개의 소나신호를 대상으로 정준상관분석을 수행하여 두 신호의 상관성을 분석하였다.

Semi-Partial Canonical Correlation Biplot

  • Lee, Bo-Hui;Choi, Yong-Seok;Shin, Sang-Min
    • 응용통계연구
    • /
    • 제25권3호
    • /
    • pp.521-529
    • /
    • 2012
  • Simple canonical correlation biplot is a graphical method to investigate two sets of variables and observations in simple canonical correlation analysis. If we consider the set of covariate variables that linearly affects two sets of variables, we can apply the partial canonical correlation biplot in partial canonical correlation analysis that removes the linear effect of the set of covariate variables on two sets of variables. On the other hand, we consider the set of covariate variables that linearly affect one set of variables but not the other. In this case, if we apply the simple or partial canonical correlation biplot, we cannot clearly interpret other two sets of variables. Therefore, in this study, we will apply the semi-partial canonical correlation analysis of Timm (2002) and remove the linear effect of the set of covariate variables on one set of variables but not the other. And we suggest the semi-partial canonical correlation biplot for interpreting the semi-partial canonical correlation analysis. In addition, we will compare shapes and shape the variabilities of the simple, partial and semi-partial canonical correlation biplots using a procrustes analysis.

An Application of Canonical Correlation Analysis Technique to Land Cover Classification of LANDSAT Images

  • Lee, Jong-Hun;Park, Min-Ho;Kim, Yong-Il
    • ETRI Journal
    • /
    • 제21권4호
    • /
    • pp.41-51
    • /
    • 1999
  • This research is an attempt to obtain more accurate land cover information from LANDSAT images. Canonical correlation analysis, which has not been widely used in the image classification community, was applied to the classification of a LANDSAT images. It was found that it is easy to select training areas on the classification using canonical correlation analysis in comparison with the maximum likelihood classifier of $ERDAS^{(R)}$ software. In other words, the selected positions of training areas hardly affect the classification results using canonical correlation analysis. when the same training areas are used, the mapping accuracy of the canonical correlation classification results compared with the ground truth data is not lower than that of the maximum likelihood classifier. The kappa analysis for the canonical correlation classifier and the maximum likelihood classifier showed that the two methods are alike in classification accuracy. However, the canonical correlation classifier has better points than the maximum likelihood classifier in classification characteristics. Therefore, the classification using canonical correlation analysis applied in this research is effective for the extraction of land cover information from LANDSAT images and will be able to be put to practical use.

  • PDF

Canonical Correlation: Permutation Tests and Regression

  • Yoo, Jae-Keun;Kim, Hee-Youn;Um, Hye-Yeon
    • Communications for Statistical Applications and Methods
    • /
    • 제19권3호
    • /
    • pp.471-478
    • /
    • 2012
  • In this paper, we present a permutation test to select the number of pairs of canonical variates in canonical correlation analysis. The existing chi-squared test is known to be limited to normality in use. We compare the existing test with the proposed permutation test and study their asymptotic behaviors through numerical studies. In addition, we connect canonical correlation analysis to regression and we we show that certain inferences in regression can be done through canonical correlation analysis. A regression analysis of real data through canonical correlation analysis is illustrated.

UNIFYING STATIONARY EQUATIONS FOR GENERALIZED CANONICAL CORRELATION ANALYSIS

  • Kang Hyun-Cheol;Kim Kee-Young
    • Journal of the Korean Statistical Society
    • /
    • 제35권2호
    • /
    • pp.143-156
    • /
    • 2006
  • In the present paper, various solutions for generalized canonical correlation analysis (GCCA) are considered depending on the criteria and constraints. For the comparisons of some characteristics of the solutions, we provide with certain unifying stationary equations which might to also useful to obtain various generalized canonical correlation analysis solutions. In addition, we suggest an approach for the generalized canonical correlation analysis by exploiting the concept of maximum eccentricity originally de-signed to test the internal independence structure. The solutions, including new one, are compared through unifying stationary equations and by using some numerical illustrations. A type of iterative procedure for the GCCA solutions is suggested and some numerical examples are provided to illustrate several GCCA methods.

디지털 오디오 위조검출을 위한 마이크로폰 타입 인식 (Microphone Type Classification for Digital Audio Forgery Detection)

  • 석종원
    • 한국멀티미디어학회논문지
    • /
    • 제18권3호
    • /
    • pp.323-329
    • /
    • 2015
  • In this paper we applied pattern recognition approach to detect audio forgery. Classification of the microphone types and models can help determining the authenticity of the recordings. Canonical correlation analysis was applied to extract feature for microphone classification. We utilized the linear dependence between two near-silence regions. To utilize the advantage of multi-feature based canonical correlation analysis, we selected three commonly used features to capture the temporal and spectral characteristics. Using three different microphones, we tested the usefulness of multi-feature based characteristics of canonical correlation analysis and compared the results with single feature based method. The performance of classification rate was carried out using the backpropagation neural network. Experimental results show the promise of canonical correlation features for microphone classification.

편정준상관 행렬도 (Partial Canonical Correlation Biplot)

  • 염아림;최용석
    • 응용통계연구
    • /
    • 제24권3호
    • /
    • pp.559-566
    • /
    • 2011
  • 행렬도는 이원표 자료행렬의 행과 열을 탐색하기에 유용한 그래프적 방법이다. 특히, 정준상관 행렬도는 정준상관분석의 결과를 이용하여 두 변수군과 개체간의 관계를 기하적으로 살펴볼 수 있다. 그 반면에 자료의 성격에 따라 세개 이상의 변수군이 존재하는 경우에는 정준상관분석의 개념에서 확장한 일반화 정준상관분석을 이용하여 일반화 정준상관 행렬도를 고려할 수 있다. 그러나 자료의 성격에 따라 두 변수군 외에 이들 두 변수군에 선형적 영향을 미치는 공변량변수로 이루어진 다른 한 변수군이 존재하는 경우에, 일반화 정준상관 행렬도를 적용한다면 공변량변수군의 영향력 때문에 주 관심인 두 변수군에 대하여 잘못 해석할 수 있다. 따라서 본 연구에서는 Rao (1969)의 공변량 변수군의 영향력을 제거한 편정준상관분석을 살펴보고, 이를 기하적으로 해석하기 위한 편정준상관 행렬도를 제안한다.