CS519: Research Methods in Computer Science

How do you choose a research project? How do you write a thesis proposal or an NSF proposal? What research methods can be employed to carry out the research? How should the research be written up and presented? This course is intended to answer these questions by giving students practice and experience with these tasks. The course will present a model of the "research lifecycle" that provides an organizing framework for research problems and research methods. Students will practice critical reading and writing of research papers. Students will also write and review research proposals. This course is highly recommended for PhD students who are preparing their dissertation proposals, but anyone interested in a research career is encouraged to enroll.

Prerequisite: Graduate standing in Computer Science.

• CS519 newsgroup. Important announcements concerning assignments, due dates, etc. are posted to the orst.cs.519 newsgroup.
• Syllabus.
• Part 1: The research life cycle (postscript slides).
• Possible papers for "Assessment Reports" in Concept Learning:
  • L. Breiman (1996) Bagging Predictors.
  • Yoav Freund and Robert E. Schapire. (1996). Experiments with a new boosting algorithm. In Machine Learning: Proceedings of the Thirteenth International Conference, 1996.
  • Ross Quinlan. (1996). "Boosting, Bagging, and C4.5. Proceedings of AAAI, 1996.
• Part 2: Research Methods (postscript slides). This concentrates primarily on relevant statistical methods.
• Files for the data analysis assignment:
  • p1
  • p2
  • p5
  • p6
• Part 3: Professional Career Issues (postscript slides). Writing and reviewing grant proposals, giving talks, ethical issues.
• Dietterich's rules of English.
• Marie des Jardins' paper on How to be a Good Grad Student/Advisor.
• What Every New Grad Student Should Know" and other useful information at Indiana University.
• CS Graduate School information at UMBC.