Project Proposals due: February 27, Friday 11:59pm
Final Reports due: March 20, Friday 11:59pm
Projects are intended to give students the opportunity to explore ideas or directions in Machine Learning that we cover in class (e.g., decision three learning, neural network, Bayesian learning, SVM, multiple kernel learning, inductive logic programming, reinforcement learning), to design interesting learning systems and consider ideas from class in more depth, or to survey current machine learning research topics or application areas (e.g., data mining, bioinformatics, Web, social networks).
Project work will include a demo for graduate students and a 8 page research paper for all, submitted as final reports.
Project proposals are to be 5 pages in length and will include a definition of the area, the particular problem to be considered in that area, and a list of several references that will support the project. The Machine Learning Information Site has a wealth of connections for finding a project.
Undergraduate projects are not expected to include an implementation effort. A literature survey (on several articles or book chapters) and a thoughtful analysis/synthesis/discussion is fine. Undergraduate students can do implementation and demo for extra credit.
Graduate students should plan on either extending software from class, developing new software, or downloading and experimenting with existing software. Graduate students will show a demo to the instructor in the final week. Projects will be expected to be individual work.
Your final paper and demo should clarify the following issues: