Schedule
- 11 of September 14:15 @ Liivi 2 - 511: Kick-off seminar
- ?? of September 14:00 Estonian time: Topic confirmation
- ?? of September 12:00 Estoninan Time: Deadline for the research plan
- ?? of October 17:00 Estoninan Time: Last chance to unregister from the course
- ?? of November 23:59 Estonian Time: Deadline for the first draft to be submitted for reviewing
- ?? of December 23:59 Estonian time: Deadline for reviews
- ?? of January 23:59 Estonian time: Final version of the project report
Under construction
Supporting seminars
- ?? September 14:15 @ Liivi 2 - 511: How to write a research plan
- ?? November 14:15 @ Liivi 2 - 511: How to do presentations
- ?? December ??: How to write research reports
Detailed instructions
Topic confirmation
To confirm a topic send me (swen@math.ut.ee) an email where you tell me the topic and a supervisor.
The supervisor must confirm your choice. If she or he does not confirm you fail the course.
Supporting lectures
There will be three lectures that help you pass the course.
- How to write research plan (Kaur Alasoo, 20th of September 14:15 @Liivi 511)
- How to present your results (Dmytro Fishman 29th November 14:15 @ Liivi 2 - 511)
- How to write research report
These will be held together with Bioinformatics seminar and they are semi-mandatory. You have to have a compelling reason to skip the seminar and this should be agreed with Sven Laur before the event.
Research plan
Research plan is a short plan that describes:
- What is the problem you are trying to solve?
- Why is this problem worth solving at all?
- What is the original data you are analysing and how to you acquire it?
- The data must be physically present or there must be a sound plan to get it the first month
- Do you need to label or assess individual data items manually?
- There must be a clear and feasible plan to get desired labelling
- What do you plan to do with the data?
- What are planned preprocessing steps?
- What kind of statistical methods you plan to apply?
- What are the desired end results?
- How do you assess validity of obtained results?
- What is the corresponding evaluation procedure?
- What is the baseline goodness you must exceed?
- What do you do if you fail to meet the evaluation goals?
- How do you determine what went wrong?
Research plan defence
After you have submitted the research plan you have to defend it. You have to present it to 2-3 researchers who will give feedback and make sure that you plan is more or less realistic and useful. If you fail the defence you fail the course.
First draft
This draft should be 80% feature complete so that your fellow students could review it and give a useful feedback. It should be 6-10 page PDF file that contains all your findings and initial results of your experiments.
Final report
The final report should be written using The offical ACM Proceedings Format. The length of the report should range from 4 to 10 pages. If you have really large illustrations then the report can be longer. Please add a subtitle "Limited access" if you report is not public by default and mention it also in the letter. Note that document templates exist for LaTeX and Word! For Lyx fans there is also a Lyx template. Please submit the final report as a PDF file. If your work is confidential then add subtitle restricted access so that I could not guess it. If you worry that Google indexes your work, then do one of the following:
- Make the PDF binary (remove easily extractable text)
- Print the work out and scan it in without OCR
All non-confidential works will be made downloadable in the course page