CSE 590x assignment 1: write a research proposal

Write a research proposal. This should not be a narrative of work you have done in the past, but a proposal about one specific project that you could accomplish in a year or two. If your project is part of some larger effort, focus on your part. Your proposal should lay out the need for the project, the benefits of a solution, how your approach differs from previous work, and why your approach may succeed. If your proposal answers the questions in the Heilmeier Catechism, it is a good proposal. However, for this 2-page proposal, you should omit the cost and project length.

This proposal should be up to 2 pages long. The page limit includes the bibliography, figures, etc. The proposal should start with a descriptive title (not "research proposal") and your name. The proposal should be aimed at a computer scientist who does not work in your field, such as another graduate student or the instructor of the class.

Please bring 3 copies to class.

Here is an example proposal. Your proposal should have a descriptive title and your name, though.

If you cannot think of a research project that you would like to perform, then write a proposal for your current research project. If you are not currently working on a project, then write a proposal for a project you have done in the past. If you understand your goals or current project, this should be quick to write. If you do not understand them, then this assignment will yield great benefits for your work, beyond the class.

This is not necessarily an easy assignment. (Also, the workload will get lighter in the class! After the first 3 weeks, there will rarely be homework.) However, it is important because it will help you develop essential research skills. If you are not already having conversations with your advisor along exactly the same lines as this assignment, you should be!

If you submit your assignment at least 24 hours early, the instructor will give you individual feedback.

Fellowships

Are you in the midst of applying for a fellowship?

• If so, then you may use the same document, so long as it fits in the guidelines above. Optionally, you may submit your personal statement for feedback as well.
• If not, you should be!

  If you are a US citizen or permanent resident (green card), then you should definitely apply for a NSF Graduate Research Fellowship. Here are some reasons why.

  • The fellowship looks great on your c.v. It makes you more attractive to employers (whether academic or non-academic).
  • It reduces the funding load on your advisor, giving your advisor more time to spend with you.
  • You are only allowed to apply for it once, in your first year of graduate school. If you don't apply for it this year, you have lost your opportunity.
  • Applying helps all computer scientists. The reason is that NSF grants a number of fellowships based on the number of applicants. If twice as many physics grad students as CS students apply, then NSF will give twice as many fellowships in physics as in CS. Thus, even if you don't think you will win the fellowship (and you might be wrong about that!), you should still apply.