Here is a collection of advice from different people about how to do good research and write good papers.
From Simon Peyton Jones
- Don’t wait, write.
- State your contributions
- Giving credit to others does not diminish the credit you get from your paper
- Introduce the problem, and your idea, using examples and only then present the general case. Explain it as if you were speaking to someone using a whiteboard
From Ross Girshick
“Show the simplest form of your idea.”
1. Start from a solid baseline, 2. Apply your idea to it, 3. Perform ablations under simple settings.
“What did I learn from your paper?”
A paper should be about a single focused idea or question. “Idea” usually means method; What should I learn? -. Under what conditions does it work? When does it work? If the idea has multiple components, which are the most important? Which implementation details are important?
“Ablations: One Table, One Message”
Example: Mask R-CNN paper. Many tables, one message per table.
“Support all of your claims”
All claims should be supported. By citation or by experiment.
Otherwise qualify the statement. “Intuitively, increasing X is important for Y..” This statement is your intuition (not fact), the reviewer may disagree.
“Increasing X may lead to improved Y…” Expresses uncertainty or that some conditions may apply.
From Joseph Paul Cohen
I have common sayings to guide research projects. If you work with me you will probably hear one of these.
+ Always work on the MVP (Minimum Viable Paper)
+ When writing a paper start with the first approximation, an outline, and refine it iteratively.
+ Test your code by breaking it