Department of Statistics

STAT302: Graphics, Multivariate Methods and Data Mining

Lecturer in charge

Dr Tania Prvan

Availability

E2 - Evening; Offered in the second half-year

Unit Outline

Websites

Handbook entry

Coursework Information (Login required)

Description

This unit consists of two modules concerned with the structure of multivariate data, one analytical, the other graphical.

The graphical module provides an introduction to a selection of topics related to new computer-based displays of multivariate data. Topics include: table presentation, principles of good graphical design, the scatterplot matrix, and dot charts for one and two-way classified data.

The multivariate analysis module provides an introduction to selected topics in multivariate analysis; namely, cluster analysis, principal components, multidimensional scaling and discriminant analysis. Knowledge of simple matrix algebra, although not essential, would be very helpful in understanding and working through these topics. Extensive use will be made of statistical packages to illustrate the concepts in lectures and tutorials.