Assignments in Data Mining

Assignments in Data Mining

Erik Zeitler (, office: P1320)

Gyozo Gidofalvi (, office: P1316)

October 22, 2008

1.Background

The lab course in Data Mining is examined using both a hand-in of the solutions and an oral examination of each assignment. The reason for having the oral examination is to make sure everybody understand the theory behind the solutions. The assignments cover a large part of the course material. Active participation in the assignments will take you a good way towards passing the final exam.

2.Preliminaries

• The work on the assignments should be performed in groups of 2 – 3 persons. Groups of one person are not allowed.
• Discussions between groups are recommended, but collaborating on the actual solutions is considered cheating and will be reported. If you are in doubt where the line is drawn, please ask the instructors in order to avoid any unfortunate misunderstandings.
• All implementations should be runnable and compilable on Windows XP, Linux (we use kernel 2.6.9, libc-2.3.4, gcc-3.4.6), or on the Solaris (Unix) system at the IT department.If you develop the solutions elsewhere, please make sure that the above is true before you hand in the solution.
• Please read the instructions carefully and follow them, especially in the cases where a certain output is demanded. This will speed things up considerably in the examination.

3.Examination

• Before each deadline, lists will be posted outside the instructor’s office where each group should sign up for a time to perform the examination. The times should be booked at least one day before the deadline.
• Make sure that you send your solution by e-mail to the instructor at least 24 hours before the examination. That will save a lot of time during the examination.
• Each person should bring a printout of the form with all the questions answered. There is one form per assignment. If you get a pass or completion grade, we will write our comments on your form.
• The examination starts with a verification of your solution. Since the time is limited, itis very important that you follow the assignment instructions exactly with respect to the formatof the answers the solution should give. If we have to spend a large proportion of the examination on technicalities because your solution does not comply, we will not have enough time to talk about the theory of your solution. Consequently, we will have to fail all members of the group then.
• When the correctness of your solution is established, the oral examination begins.The instructorwill ask questions to each of the group members to establish that all members haveparticipated in and understood all parts of the assignments. The questions will be about your solution, but also about the theory that your solution builds upon. In general, you should notonly have solved the problem, but also have understood how your solution works.
• At the end of the examination, you will get one of the grades pass, fail, or completion. This isbased both on the solution and on the oral examination. It is possible thatone group member pass and the others fail.
• The completion grade is given if the solution is incomplete in a small way or if you have someminor lack of knowledge in some area. The instructor will tell you on a case-by-case basis what you need to do in order to pass a completion.
• If you fail the assignment, you have to redo the examination later. There will beadditional examination opportunities at the end of the course. Dates will be posted on the course home page.

4.Questions

If you have any question, you can visit us during our office hours, or send an e-mail. As a general remark, do not hesitate to ask us about any problems during the course.

Good luck!