Fall 2011, Thursdays, Milner 216, 4:00-4:50 PM

Date: December 1

"  Two Approaches to Binary Classification of Data "
  Jennifer Webster  


As researchers, we come across many situations where we wish to classify items based on some indirect measurements. For example, we predict passing or failing a course based on the grades from the first half of the semester. Even in this simple example, we can see that incorporating more assignments and tests makes our predictions more accurate than analyzing a single instance. This talk explores two-category classification using sensor array data through two approaches - shape optimization and a probabilistic method- by examining a border security problem.