NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/comp/ai/4322

Text File | 1992-11-17 | 3.4 KB | 68 lines

Newsgroups: comp.ai
Path: sparky!uunet!ukma!darwin.sura.net!mojo.eng.umd.edu!clin
From: [e-mail] (Charles Lin)
Subject: Re: How to pick a grad school [excerpt from Stephen Gould]
Message-ID: <[e-mail]>
Date: Tue, 17 Nov 92 22:24:35 GMT
Organization: College of Engineering, Maryversity von Uniland, College Park
Sender: [e-mail] (Charles C. Lin)
References: <[e-mail]> <srt.721931824@sun-marino> <[e-mail]>
Originator: [e-mail]
Lines: 55


In article <[e-mail]>, [e-mail] (Lou Steinberg) writes:
>In my experience (meaning Stanford some (too many :-) years ago and
>Rutgers since then), you are not really choosing an advisor when you
>choose a school.  Rather, you are choosing a small pool from which
>your advisor will be drawn.  There have been several cases at Rutgers
>where students began working with one advisor, found they could not
>get along with him (or he with them), and switched to another.
>Students have switched from me to my colleagues and from my colleagues
>to me.  In many cases, students who almost certainly would not have
>finished had they not switched have gone on to do theses ranging from
>creditable to excellent.  In fact, thinking about specific cases, I can
>think of more where this has happened than where it has not.

    I think I mentioned something like this in a previous article.
In any case, choosing a grad school with a group of faculty in
some broad general area (like AI) and perhaps even in more specified
areas within AI (say, distributed or parallel AI) not only helps
because there's some freedom to choose from a group, but also because
the department has some commitment to research in that area.

>
>This argues for choosing a school with several faculty members you
>think you might work with.  It also argues for being flexible about
>thesis topic and, to some extent, area.  After all, what are the
>chances that 15 years from now you will be working on the same narrow
>area as your thesis?  (It has been common in my experience that a
>student doesn't even get through a *thesis*, let alone a career,
>without some substantial change in focus.)

   I wonder how many students come in with a thesis idea ahead of
time.  Considering that grad school may be the first time namy
students have had an opportunity to do research, most of them
probably do not have a clear idea of what research to do.  In any
case, ultimately, they will have to work with some professor or
advisor, and the research topic selected will most likely have to
conform to the interests of the professor.   It is rather uncommon
for someone to come up with a wildly original topic as a graduate
student and work on this idea outside the interest of an advisor.
Most students seem to (in the US anyway) look at what work some
prof. is doing, see if they like the work, and like working
for that faculty member, and then join (assuming he or she is
accepted to the group) in.   After all, it seems as if some students
select grad schools, not by their research strengths in a *specific*
area, but by the reputation of the institution.   Therefore, all
the top students seem to want to go to Stanford or MIT of CMU or
wherever, and only when they look deeper do they find places like
Rutgers or U. of Oregon, etc. which may have just as good a program.

--
Charles Lin
[e-mail]