General description

When several variables are measured on each experimental unit, the result is a multivariate statistical data set. Multivariate methods are needed when one wishes to analyze several variables simultaneously. Matrix calculus is used as a mathematical tool.

Learning outcomes

The student should learn to recognize the most typical situations requiring multivariate analysis and to make the most common analyses using some statistical software.

Contents

Graphical inspection of multivariate data, properties of the multivariate normal distribution, multivariate tests, principal components, factor analysis, discrimination and classification, clustering.

Teaching language

Finnish

Evaluation

Numeric 1-5.

Recommended year of study

2. year spring
3. year autumn
3. year spring

Study materials

1. Everitt B., Dunn, G., Applied multivariate data analysis. Arnold 2001.
2. Everitt, G., An R and S-PLUS companion to multivariate analysis. Springer 2007.
3. Johnson, R. A., Wichern, D. W., Applied multivariate statistical analysis. Prentice-Hall 2002.
4. Mardia, K. V., Kent, J. T., Bibby, J. M., Multivariate analysis. Academic Press 1979.
5. Mustonen, S., Tilastolliset monimuuttujamenetelmät. Helsingin yliopisto 1995.
6. Sharma, S., Applied multivariate techniques. Wiley 1996.
7. Srivastava, M. S., Methods of multivariate statistics. Wiley-Interscience 2002.

Belongs to following study modules