Projects
An open-ended term project is a major part of the course. In the term project, you will investigate some interesting aspect of machine learning or apply machine learning to a problem that interests you. For advice with the project topics, please contact the instructor.
Team
A team of maximum two persons is allowed.
Schedule
Project proposal will be due on Sep 20 (one page). There will be a mid-term progress report (maximum 4 pages) due on Nov 5. For mid-term project presentation schedule (maximum 5 min presentation, maximum 6 slides), see below. (The midterm report and presentation will be separately graded.) The final project report will be due at the end of the semester (maximum 8 pages) on Dec 15. Final presentation on Dec 15th, 6:30-9:30pm.
Report format
It is suggested that the proposals/reports be in a single column format, with atleast 11 font size. It is not necessary, you can use this format for the reports.
Fall 2009 projects
6:30pm-7:55pm: Robotics/Vision track
Combine models for scene understanding. Adarsh Kowdle, Congcong Li.
Fleet control of semi-autonomous robots using a written gesture based interface. Danelle Shah, Joe Schneider.
Using Social data to recommend images. Kent Sutherland.
Estimate egomotion and detect obstacles simultaneously for moving aerial robot. Tung Sing Leung.
Object Style Recognition. Fang Liu, Daniel C. Hauagge.
Landmark Recognition. Henry Shu.
Function-based classification of human hand tools. Guilherme Pinto.
Indoor Autonomous Helicopter Navigation. Yu-hsin Chen.
Chemotaxis of Dictyostelium. Diana Chang.
8pm-9:30pm: Learning / Theory / Applications track
Using machine learning to prove theorems in NuPRL. Jean-Baptiste Jeannin.
Belief Updates in Games with Unawareness. Samantha Leung.
Predicting Energy Usage. Yun Jiang.
Prefetching in Fabric using Learning. Evan Danaher.
Boosting. Mark Anthony Carty
Learning triples abstraction from text. Ruben Sipos.
Predicting the outcome of competitive games based on prior player game history. Jason Hardy.
Learning to Balance an Inverted Pendulum. Sarabjeet Singh.
A Variety of Confidence Scores for kNN Classification. Ramu Nachiappan.