[Work Log] Talk w/ Kobus Re: Dissertation Proposal

January 13, 2014

Met w/ Kobus today to discuss the proposal for the first time.


Key idea: Branching GP.

Close 2nd: Model selection

Other stuff like high-throughput phenotyping, biological modeling, multi-view stereo, etc. take a back seat (but should be mentioned)

Key sections

The structure isn't rigid, but the three main sections that need to be strong are: (1) Literature review, (2) current progress, (3) future work.

Didn't ask about "Assumptions and Limitations". Kobus mentioned that things like that can protect you, but they don't come up very often. Seems not critical, use best judgement.

"Broader impacts" or "Importance of Topic" section could be a heading to what would otherwise just fall into the introduction.

It seems the "Outcomes" section is probably best weaved into the "proposed work".

It seems listing the dissertation chapters isn't strictly necessarilly, either.

Future work

Full automatic system
Automatic camera detection
Apply to another domain (can be vague; can change later)

Less likely:
    leafs, flowers, etc.  Stick to GP's


Don't stress about correctness.  This is a proposal, i.e. *pre* investigation.  
Let's wait until the end of the week to decide if we need to reschedule.
Posted by Kyle Simek
blog comments powered by Disqus