Multivariate analysis skills have been commonly recognised as part of the key requisites for analytics analysts. The complexity of most phenomena in the real world requires an investigator to collect and analyse observations on many different variables instead of a single variable. The desire for statistical techniques to elicit information from multivariate dimensional data thus becomes essential and crucial for data analysts.

The objective of the course is to introduce several useful multivariate techniques, making strong use of illustrative examples and matrix mathematics. The course will start with the extensions of univariate techniques to multivariate frameworks, such as the multivariate normal distribution, confidence ellipse estimation, hypothesis testing, simultaneous confidence intervals and Bonferroni confidence intervals. The course will also cover the techniques unique to the multivariate setting such as principal component analysis, factor analysis, discrimination, classification and clustering analysis.

Skills will be developed with SAS, a leading statistical analysis software package used in industry.