At Columbia, PhD defenses commonly are not public, although CS usually enables students audience. Defenses contain four parts: first, the candidate introduces themselves, then presents a listing of the work they do, interrupted and adopted by questions in the committee. Finally, the committee meets privately to go over the presentation and dissertation.
While the majority of the committee may have read much of your thesis, you can’t think that everybody has read every chapter.
The committee needs so that you can assess impact and depth. Usually, the committee has an idea of the prior to the defense, but regardless of the student know to create this assessment simpler, possibly just through emphasis, could make the defense go a lot more easily.
Generally, the entire defense won’t take greater than two hrs, but must take significantly a shorter period. Area of the challenge of the defense would be to convince the committee that you could summarize information of the work in an exceedingly short time.
Your presentation (and thesis) must address the next:
What’s the problem you’re studying?
Why do important ?
What results have you ever achieved?
Some committee people may wish to determine if the whole shebang continues to be printed and how and where it had been received. For instance, for those who have written software, indicate where it’s getting used, because of follow-on work or perhaps in some production or test atmosphere.
Have a summary of your thesis-related publications like a slide. Indicate any awards that the paper might have received. For most of us, it’s simpler to list out some recognition than “brag” about this personally.
For those who have presented your projects inside a conference or at job talks, make sure to anticipate and address the most typical questions requested there.
The committee ought to be handed a duplicate of the slides.
Be ready to briefly summarize your background (undergraduate degree, how lengthy in the college, etc.)
A maximum of 3 slides, plus “supportInch slides with a lot more material in situation of questions. The best way of creating your committee people mad would be to come unprepared with a collection of 8 slides after which madly skip through them.
Number your slides, especially if your committee people is related in via speakerphone. Think about using some type of remote presentation software.
List your contributions early.
When presenting your contributions, make sure to use “I” and never “we” so the committee knows what facets of the job where yours, and that have been group projects.
Keep discussions of related work very brief, but be ready to respond to questions from the “so how exactly does this vary from so-and-so’s work” succinctly.
You won’t be requested to demonstrate results again.
Be ready to support any comparative statement with details, particularly statements like “works more effectivelyInch, “faster”, “scalable” or “optimal”. If you’re presenting a protocol, how can you tell it works properly?
For those who have multiple parts inside your dissertation, talk to the committee in advance whether it seems sensible to omit a number of them for that presentation.
It is best to concentrate deeply on one area then to operate on several topics, because both versions is went after to some moderate depth.
Systems work should be along with implementation and some type of statistical comparitive analysis to show the enhancements from existing or alternate approaches.
Your thesis requires a one page executive summary that the layperson will be able to understand. Test: create it for a family member you have without an engineering degree.
It is possible simply to defend a PhD thesis just once your defense is a special event, so consider dressing appropriately, a minimum of business casual, however a suit isn’t inappropriate.
It’s customary to supply refreshments for that audience, for example coffee, bagels, cookies and fruit, with respect to the time.
The Function of PhD Committee People
Committee people (should) browse the draft thesis (and supply feedback). Clearly, students appreciate an in-depth studying, but it’s common for committee people to pay attention to chapters nearest for their expertise. Studying depths varies – some provide line edits, others just suggest bigger issues that needs to be addressed (“Your related work section in Chapter 1 is sparse and leads to 5.”). Although this is most likely not the area to point out “do another year of research”, filling out gaps is alright and I’d rather postpone a defense with a month as needed. Prior to the committee will get the thesis, I have done an initial or sometimes second studying, however the whole idea from the committee would be to keep your consultant honest (and complement their understanding or taste).
Committee people attend the PhD defense, usually personally. Typically, this lasts about 1 hour 3 minutes. Take notes on any editorial enhancements (e.g. “make obvious the throughput graph is measured in gallons/minute”). Election around the outcome and sign the shape.
When the student is offered some changes to apply, the consultant asks students to detail about how they implemented the alterations, much like how a writer may react to reviewer comments for any journal. The committee informally signs off, or otherwise, on these changes. There’s you don’t need to re-browse the thesis.
Before you submit your draft towards the committee, make sure to verify you have done the next checks:
Look for missing chapter or figure references
Section, Chapter, Figure are capitalized
All references converted from  to [1,,3]
Consistent capital in captions
Verify growth of all abbreviations initially instance
Avoid “tremendous”, “huge” along with other similar adjectives
Finish to finish -> finish-to-finish
Check references for capital of abbreviations and missing data for example page figures.
(Contributions by Erectile dysfunction Coffman, Jonathan Rosenberg and Sal Stolfo.)