Development of Some Multivariate Statistical Inference Procedures for Missing and High Dimensional Data and Its Application

点击翻译按钮获取中文摘要

We obtained some results for the tests of mean vectors when the data have missing observations. In particular, we gave the test procedures for mean vectors using the likelihood ratio test statistic and Hotelling type test statistic for the one sample and two sample problems when the data set has two-step monotone missing data. We also gave the simultaneous confidence intervals for any and all linear compounds of the mean. Further, the accuracy of the approximation is investigated by Monte Carlo simulation. Testing specified values for the mean vector and the covariance matrix with a two-step monotone missing data were considered and derived the likelihood ratio test statistic and some results. As for the study of high dimensional data, we obtained some results for the problem of discriminant analysis and testing for multivariate normality and so on.