Course Detail
Units:
4.0
Course Components:
Laboratory
Lecture
Enrollment Information
Enrollment Requirement:
Prerequisites: "C" or better in (GEOG 4140 OR GEOG 5140).
Description
Graduate students should enroll in GEOG 6160 and will be held to higher standards and/or more work. Geographical datasets are increasingly large and complex, and may consist of observations in both space and time. These characteristics limit the use of classical statistical approaches, and a number of geocomputational methods have been proposed to supplement or replace these. This class will explore two key areas of geocomputation. The first part of the semester will examine the use of machine learning techniques with large spatial datasets. This will be followed by a look at methods used to analyze and model spatio-temporal data and dynamic spatial systems. Techniques to be covered will include: neural networks, tree-based methods, maximum entrophy, grid-based and individual-based modeling. The class will consist of both a lecture component, to introduce the methods, and a lab component to allow hands-on experience. Students will be expected to complete a short research project, demonstrating the use of one or more of these methods.