BEE 4310 - Multivariate Statistics for Environmental Applications

Spring. 3 credits. Letter grades only.

Prerequisite: either multivariable calculus, linear algebra, or introductory statistics or permission of instructor. Co-meets with BEE 6310.

S. Steinschneider.

This class provides an introduction to relatively simple but powerful multivariate statistical techniques needed to analyze and model complex datasets frequently encountered in the environmental sciences. Emphasis is given to developing the mathematical foundation of these methods to foster a deeper understanding of the benefits and limitations of different approaches. The goal is to provide students in the applied environmental sciences with a toolbox of methods not taught in more introductory statistical courses, but also to ensure that students can use these methods in their own work without viewing them as a “black box”. The course only assumes a limited knowledge of calculus, linear algebra, and statistics, and will provide a review of the mathematical concepts needed to understand the multivariate techniques presented. Applications will be presented primarily from the geophysical and ecological sciences, but the theory will be applicable to other environmental fields. Upon completion of the course, students will be able to use the multivariate techniques presented in the course to understand and evaluate environmental problems of interest in their respective domains.

Outcome 1: Students will be able to explain the utility of multivariate methods for inference, risk analysis, data compression and signal extraction, classification, and clustering in high dimensions.

Outcome 2: Students will be able to conduct multivariate analyses in modern statistical programming environments.