CU.578.02 Empirical Methods in Machine Learning Rich Caruana & Alex Niculescu 12/05/02 Final Exam: - Wed Dec 18 9:00-11:30am Phillips 203 - Open Books, Open Notes - No computers or wireless devices - Exam covers entire course with emphasis on recent material Optional 4th Make-Up Assignment: - For those who did poorly on a previous assignment - New grade for worst homework is 0.75*HW4 + 0.25*HW? - Most people should not do this - Assignment is agglomerative clustering Final Project: - Predictions officially due by email Sat Dec 7 - Email predictions to both alexn at cs cornell edu *and* caruana@cs.cornell.edu - Email should have three attachments - Names of group members should be in each attached file - Will not start assessing late points until 5pm Wed Dec 11 - Extra Credit if turned in early - Final reports will not be accepted after 3pm Mon Dec 16 - Can give reports to Cindy Robinson in Upson 4146 1:30-3:30pm any day - Note: it is OK to take the average of different kinds of models, e.g., to average a neural net's predictions with a decision tree's predictions Good Luck! Thanks for making the course a success. -Rich.