Instructors: Eric Nyberg and Teruko Mitamura
Course Description: This course is primarily a project course. During the first few weeks of the course, students complete a set of readings (which are complemented by instructor presentations) and associated reading quizzes. Then the entire class will brainstorm ideas for new approaches to a current data set or leaderboard. After formulating specific hypotheses about how to extend the state of the art, the class will split into teams in order to implement and evaluate the associated experimental conditions for each hypothesis. The overall goal is to extend the state of the art performance on the given data set or leaderboard, and possibly submit a technical publication which describes the work in detail (when the results demonstrate a statistically significant improvement versus the state of the art). During each remaining week in the semester, student teams will present their incremental project progress, followed by group discussion. During Weeks 13-15, each student team will make a final presentation. A final report / paper will be due during Week 16. The final grade will be based on all the deliverables (project proposal, final presentation, final report / paper).
Grading: