Research and Graduate School
MS degree –can give you a nice boost in salary, more opportunities (e.g. project leader) –usually 2 years –2-3 courses per semester –area specialization (like networking or AI) –thesis or non-thesis –might work in a lab on a project for a professor
PhD –academic life can be a rewarding career path (independence), but also stressful (“publish or perish”) –can work in industry or academia (as professor) –usually 5-7 years –after coursework, mostly just work in lab (on dissertation and writing conference/journal papers) –might have to pass comprehensive or qualifying exams (after 1-2 years)
Getting into Grad School must take GREs (like SATs) should have good grades –competition is fierce, everybody has A’s... need recommendation letters undergrad research experience helps
Funding most grad students get paid –ok, it is a paltry salary, but positive cash flow is always a good thing –also, tuition is often waived 4 principle sources –TA – teaching assistant –GANT –non-teach, e.g. system administrator –GAR – graduate research assistant this is most coveted, but you must convince an advisor to take you into their group (limited funding positions depending on grants) –fellowships – scholarships for grads (prestigous)
A Day in the Life of a PhD student classes are harder, but take fewer at a time constantly reading research papers the goal is to write papers mentoring by their advisor is critical help out in the lab – work on funded projects, often with other grads in group group meetings go to seminars
Conferences faculty and grads try to go to conferences to.. –present their own ideas (~30 minute talks) –keep up with advancements in the field –exchange ideas with their colleagues get a paper accepted travel to interesting places for free major annual conferences: WWW, SigGraph (graphics/animation), AAAI, ICML (machine learning), ICRA (robotics), FOCS (theory) look at ICRA 2013 and accepted papers in proceedings as an example:
The Faculty Hierarchy Assistant professors – pre-tenure (7 years) –their attention is focused on getting tenure, primarily through getting grants and publications Associate professors – tenured Full professors other players: –adjunct/joint/visiting/research faculty –“post-docs” – people with PhDs but not yet faculty, ~2 year window to publish some more by working in another lab
faculty members have worked hard to get where they are, so give them respect (address them as “Dr.....”) –getting a PhD and getting tenure are arduous journeys each faculty member’s pride-and-joy is their publication list, which demonstrates their research area and productivity –see Michael Littmans’ pub list as an example – focused on applying for grants (= funding for lab and grad students) –e.g. NSF (National Science Foundation), also companies like Motorola or Ford
Publishing - this is what we do –clever ideas are not appreciated until they shared publicly 2 main venues: –conferences: 6-8 pages, 1 month review –journals: pages, 1-6 months review getting papers accepted is hard –must be a novel contribution, extending prior work –theorems or analysis must proved rigorously –algorithms must be thoroughly tested –experiments must be well-designed, with controls and statistics peer-review process –typically 3 anonymous reviewers who are experts in the area –taken seriously – reviewers will pick a paper apart and point out its flaws see JAIR (Journal of AI Research) as an example
PhD Dissertations you must carve out a niche and make a substantial contribution of new ideas ideally should have some recognition in the research community typically pages, written at the end usually multi-faceted – look at a core idea from several angles, fully explore it published papers can be turned into chapters defense – oral presentation to thesis committee (chair/advisor plus 2-3 other faculty you choose) for example PhD theses in computer science, see repository at NYU Courant Institute: