Nonlinear Multivariate AnalysisWiley, 1990 - 579 Seiten Conventions and controversies in multivariate analysis; Content analysis of MVA books; Correspondence analysis of tables of content; A shortsummary and some problems; Data analysis and statistics; Data analyticprinciples of this book; Specific problems of MVA; Definition of MVA; Some important ingredients; Epilogue; Coding of categorical data; The complete indicator matrix and its properties; Quantification; The incomplete indicator matrix; Missing data; The reversed indicator matrix; The indicator matrix for a contingency table; Grouping of categories; Grouping of variables; Epilogue; Homogeneity analysis; Homogeneity of variables; Historical preliminares of differential weighting; Maximizing homogeneity by linear a weighting: principal components analysis; Alternating least squares algorithms for linear weighting; Linear weighting for K sets of variables; More historical comments on pca; Maximizing homogeneity by nonlinear transformation: nonlinear PCA; Categorical PCA: homals; Relations between homals and linear PCA; The relationship between homals and total chi-square; An illustration: hartigan's hardware; Homals with incomplete indicator matrix; Reserved indicator matrix; Epilogue; Nonlinear principal components analysis; Metric principal components analysis; Nonmetric principal components analysis; Theory of join loss; Theory of meet loss; Geometric of meet loss; The princicals program; An example comparing multiple and single nominal treatment; An example with preference rank orders; An example with discretization of continuous variables; Epilogue; Nonlinear generalized canonical analysis; Previous work; Loss function and normalization of overals; Overals as a special case of homals: the use of interactive variables; Missing data; Algorithm construction; An example: effects of radioactivity on fish; Epilogue; Nonlinear canonical correlation analysis; Previous work; Theory; The canals program; Example 1: economic inequality and political stability; Example 2: prediction of a school achievement test; Epilogue; Asymmetric treatment of sets: some special cases, some future programs; Multiple regression and morales; Discriminant analysis and criminals; Multivariate analysis of variance and manovals; Path analysis and pathals; Partial canonical correlation analysis and partals; Some examples; Epilogue; Multidimensional scaling and correspondence analysis; Homogeneity and separation; Minimum distance analysis of homogeneous groups of objects; Correspondence analysis; Contigency and correlation; The program anacor; The program anaprof: analysis of profile frequencies; Some gauging results for binardy data; Epilogue; Models as gauges for the analysis of binardy data; Some general formulas; Monotone latent trait models; Nonmonotonic latent trait models; Order analysis of binardy matrices; Dichotomized multinormal distriibutions; Epilogue; Reflections on restrictions; The class of restrictions closed under linear transformations; Equality constraints; Other linear constraints; Zeros at specific places; Nonlinear restrictions; Epilogue; Nonlinear multivariate analysis: principles and possibilities; Join and meet; Choice of basis; Some infinite-dimensional gauges; Analysis of stochastic processes; Alternative criteria; Epilogue; The study of stability; Analytical stability; Algebraic stability; Replicationstability; Epilogue; The proof of the pudding; Multiple choice examination; Controversial issues; As years go by; Parliament survey; Crime and fear; Epilogue; Notation; Matrix algebra; Images; Hyperellipsoids; Invariant directions; Singular vectors and singular values; Eigenvectors and eigenvalues; Algebraic applications of eigenvectors and singular vectors; Optimization properties of the SVD; Generalized eigenvector problem; Applications in linear MVA; Joint plots; Cones and projection on cones. |
Inhalt
CHAPTER | 1 |
A SHORT SUMMARY AND SOME PROBLEMS | 19 |
5 | 37 |
Urheberrecht | |
30 weitere Abschnitte werden nicht angezeigt.
Häufige Begriffe und Wortgruppen
algorithm alternating least squares ANACOR ANAPROF binary bootstrap canonical analysis canonical correlation analysis canonical variables category points category quantifications chapter column complete indicator matrix compute constraints correlation matrix correspondence analysis data analysis data matrix define delta method diagonal matrix dimension dimensionality discrete discrimination measures discussed distribution eigenvectors elements equal equation example Figure formula frequency gauges gives GjYj Guttman scale HOMALS solution homogeneity analysis implies interpretation iterations largest eigenvalue Leeuw linear loss function marginals maximize meet loss methods minimize missing data monotone multinormal multiple nominal nonmetric normalization object scores optimally scaled option orthogonal orthogonal polynomials partitioned plot polynomials possible PRINCALS principal components analysis profiles programs PvdA rank numbers regression restrictions sample Section single nominal single ordinal single variables singular value decomposition splines SSQ(X statistics submatrices subspaces sum of squares Suppose techniques theory variance vector weights zero