Latent class analysis is based on the assumption that within each class the observed class indicator variables are independent of each other. We explore a new Bayesian approach that relaxes this assumption to an assumption of approximate independence. Instead of using a correlation matrix with correlations fixed to zero we use a correlation matrix where all correlations are estimated using an i...

in some statistical process control applications, the quality of the product is characterized by thecombination of both correlated variable and attributes quality characteristics. in this paper, we propose anovel control scheme based on the combination of two multi-layer perceptron neural networks forsimultaneous monitoring of mean vector as well as the covariance matrix in multivariate-attribu...

A principal component analysis based on the generalized Gini correlation index is proposed (Gini PCA). The PCA generalizes standard variance. It shown, in Gaussian case, that equivalent to PCA. also proven dimensionality reduction matrix, relies city-block distances, robust outliers. Monte Carlo simulations and an application cars data (with outliers) show robustness of provide different interp...

In this paper, we propose a new componentwise estimator of a dispersion matrix, based on a highly robust estimator of scale. The key idea is the elimination of a location estimator in the dispersion estimation procedure. The robustness properties are studied by means of the influence function and the breakdown point. Further characteristics such as asymptotic variance and efficiency are also an...

In this contribution the impact of a presumed presence or absence of the ionospheric delays on the quality of the least-squares ambiguities is analysed. The spatial correlation and the time correlation of the ionospheric delays are captured in an a priori ionospheric variance matrix. It is shown how the size and shape of the ambiguity search space is a€ected by the use of these a priori ionosph...