Everything happens somewhere and sometime. But the study of data collected over multiple times and locations requires special methods, due to the dependence structure that relates different observations. In this course, we’ll look at exploring, analyzing, and modeling this kind of information–introducing standard methods for purely time-series and purely spatial data, and moving on to methods that incorporate space and time together. Topics will include autocovariance structures, empirical orthogonal functions, and an introduction to Bayesian hierarchical modeling. We’ll use R to apply these techniques to real-world datasets.
The Class: Type: lecture
Requirements/Evaluation: project work, homework, exams, and contribution to discussion.
Extra Info: may not be taken on a pass/fail basis; not available for the fifth course option
Prerequisites: STAT 346, or permission of instructor
Enrollment Preference: Seniors and Statistics majors
Distributions: Division III; Quantative/Formal Reasoning;
Distribution Notes: This is an intensive statistics course, involving theoretical and mathematical reasoning as well as the application of mathematical ideas to data using software.