February was dedicated to paper writing. We had to run our experiments using final parameters, make figures, write the paper and submit it. Submission is not as simple as one can think.
We have set the final parameters for our model using grid search.
All experiments were run again to ensure validity of the figures. They are very important as they convey our message. Some of them were really hard to obtain as the gain brought by some methods was not obvious.
I knew that writing was one of my weaknesses. I do not write so bad, I think, but writing takes me ages. Unfortunately, we don't have much time when submitting a paper. That is why Gael helped me a lot by rewriting and reformulating a lot, especially mathematics formalism.
Another difficulty is the 8 page limit. On one hand, it forces people to summarize their message, but on the other hand, it forced us to remove some didactic explanation and to go right to the fact. Some tricks can save space though:
- inline simple equations
- use short formulations (for example, as opposed to is a good candidate to replace the heavy on one hand... on the other hand)
- paragraph should always ends at the end of a line: a one-word newline is a waste of space
Submission process is long and should not be rushed. This consists in:
- Listing your contributions (again)
- Listing closest papers
- Choosing commitee members that could handle your submission
- Chossing potential reviewers