Some advice on my syllabus
Andrea Taverna
I'm a recently graduated student of computer science and I'm going to
start the master degree course soon.
I've got good fundamentals of math, statistics, algorithmics and
operation research. I've dabbled with soft computing(neural nets) as
well.
For the next degree I'd like to specialize on these subjects,
especially operation research, and I'd need some advice on choosing
the courses for my syllabus.
There are a few compulsory courses on statistics, math/logic and I've
already selected the complementary course on operation research.
Apart from these, what other courses should I take?
I've eyed courses like logic programming (basic and SMT), simulation
techniques, soft computing, computational geometry and algebra, graph
theory...
They may be useful to learn, on the other hand I don't want my
syllabus to be disorganized.
thanks,
Andrea Taverna
Paul
Computer science is a good starting point. What follows are naturally
personal opinions (and, incidentally, my degrees are in math and
statistics, and I teach in a business school):
* Linear algebra is definitely useful, and I think graph theory would
be useful. Computational geometry I'm not so sure about. ("Abstract"
algebra -- groups, rings, fields -- I would say is not very useful,
although finite group theory occasionally pops up in integer
programming, at a fairly theoretical level.)
* Simulation is definitely a good choice.
* Depending on what the required statistics course covers, and
depending on whether you want a career in industry or in academe, you
might want one or more additional statistics courses. Academics (I'm
one) can to some extent "pick our fights", but in industry you have to
deal with the problems presented to you. In many cases they will
essentially resolve to statistical modeling, and even relatively
deterministic things like optimization models require numerical
parameters that need to be estimated.
* I'm not sure what the logic programming course would cover, but if
it does not cover Constraint Programming, I would try to find a course
that does.
* You did not mention courses on linear, integer or nonlinear
programming. The basic OR course will cover them to some extent, but
more is better, particularly with integer programming.
* If there is a course in data mining, I would recommend it.
* If you plan to go into industry, I would recommend a course in
business communications, presentation techniques etc. They likely
would be taught in the business department/college (if you have one).
A course in consulting (some business schools have such) would also be
useful. You'll have to talk to the suits at some point, both to
figure out what they really want you to do (which will not necessarily
be what they initially seem to be saying), and you will have to
persuade them on the wonderfulness of your finished product.
Cheers,
Paul
mjs
> On Sep 23, 6:40 pm, Andrea Taverna <a.t...@hotmail.it> wrote:
>
>
>
>
>
> > I'm a recently graduated student of computer science and I'm going to
> > start the master degree course soon.
> > I've got good fundamentals of math, statistics, algorithmics and
> > operation research. I've dabbled with soft computing(neural nets) as
> > well.
> > For the next degree I'd like to specialize on these subjects,
> > especially operation research, and I'd need some advice on choosing
> > the courses for mysyllabus.
>
> > There are a few compulsory courses on statistics, math/logic and I've
> > already selected the complementary course on operation research.
> > Apart from these, what other courses should I take?
> > I've eyed courses like logic programming (basic and SMT), simulation
> > techniques, soft computing, computational geometry and algebra, graph
> > theory...
> > They may be useful to learn, on the other hand I don't want my
Just encountered this post while browsing the group.
The INFORMS Computing Society has a report on a proposed curriculum in
computational operations research. You can view it here:
http://computing.society.informs.org/pdf/ICSed.pdf.
This is a proposal for an undergraduate degree that would prepare a
student for a computational OR graduate program. However, there might
be some inspiration within.