Life as a Grad Student
Links:
- How to Write a Master's Thesis in Computer Science
- Advice for Beginning Your PhD Career
- How to Read a Research Paper
Presentations
Things to keep in mind when writing presentations:
- Numbering slides is helpful, especially "x out of y" format
- Breadcrumbs beat outline slides sprinkled through the presentation
- Keynote lets you see two slides at once - the current and the next. This is useful for transitions
Paper Writing
Links:
Vadim's Summary of a good IJCAI paper structure:
- Important task X
- Existing method Y is inadequate for X
- New method Z proposed here
- Z is probably better
- Intuition gained
Things to keep in mind when writing papers, that have occured to me while reading them:
- Abstract: it's nice when the contribution of the paper is mentioned in the first few sentences. CiteSeer only displays the first part, and it's frustrating when that doesn't tell you what the paper is about
- Writing: there must be something you can do about "X is important" and "Y is useful." How do you avoid empty sentences introducing sections?
- Structure: not a big fan of "In Section 2 we, . . . " style, though everyone does it. Does anyone require it?
Other points:
- Dan - put a list of symbols in your thesis, so it's easy to refer back
Reviewing Papers
Minor points:- The authors don't get to see the numerical score, so state clearly whether you are accepting or rejecting