discipline

MSC in Mathematics, Specializations

subject

Multivariate statistics

lecturers

György Michaletzky. prof., Dept. of Probability Theory and Statistics, Institute of Mathematics

credits

4

period

 

curriculum

Parameter estimation for the multivariate normal distribution.

Matrix valued distributions.

Normal linear model.

Analysis of variances. Canonical correlations. Principal component analysis. Factor analysis.

Methods of classification. Bayes decision, linear and piecewise linear separation, nearest-neighbor method, cluster analysis.

Linear and nonlinear regresion. Ridge and James-Stein estimators. Stepwise and robust regression.

Analysis of contingency tables. The loglinear model. Method of minimal discriminating information. Multivariate scaling. Relaxing the assumption of normality, nonparametric and robust methods in multivariate statistics.

literature

 

form of tuition

Lectures

mode of assessment

oral exam