Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Graph Theory and Astrophysics

  1. Aug 6, 2010 #1
    Hi all,

    This is my first post to this forum, so allow me to introduce myself. I am a 28 year old person who is about to start a Phd in Machine learning with focus on graph theory. Now from my preliminary research, it seems that theory developed has a lot of use in computational biology, computational genetics etc. What I am interested in impact of graph theory in Astrophysics ? Can I shift to research in Astrophysics with a degree in Computer Science ?

    Finally, does it really matter which university I did my Phd in ? Finally, would it better, or advisable to start a undergrad(postgrad/phd ?) degree in Physics at this point in my career.

    PS: I am in Indian in Europe, so lot of places of funding is blocked for me :)
  2. jcsd
  3. Aug 6, 2010 #2
    Yes, there is. I think that this field is only going to get much bigger as well. We're at a topical time for graph theory in biology - it remains to be seen how successful it is actually going to be. We should know in a few years, so your PhD will be timely if nothing else!

    I suppose at a stretch you could look for computing jobs in physics departments. I have worked in physics departments where they have programmers working on physics problems, that didn't really have any physics background. I certainly wouldn't count on it.

    It matters that you get a good supervisor, not so much what university you come from.

    You're about to start a PhD in graph theory... Why did you accept this PhD position if you're considering taking another undergrad? You shouldn't start a PhD programme without being absolutely sure you want to do it.
  4. Aug 6, 2010 #3
    First of all, I thank you for your reply. It seems I have not been clear on my intentions, I apologise for that.

    I want to cross over to physics (quantum/astro), so I was thinking if there are applications of graph theory in these areas ? If there are, I would like to customise my research towards these areas. To this end, would I need to get another degree in physics ?
  5. Aug 6, 2010 #4
    I haven't worked with physics in a graph theory way before really, so I'm not sure about the research that exists out there - but to me it seems like condensed matter physics (which is a huge topic) is likely to have quite a few applications for graph theory. Other than that, graph theory is useful for exploring large data sets, you might find that you could try to market yourself as a data miner in physics.

    That said, if you want to go in to physics - a PhD in graph theory isn't really the way to go. You should try to do something directly related to physics. That might involve a second undergraduate degree, which is probably not ideal- it depends what you actually want to do. What I had tried to hint at in my previous reply is that physics has lots of uses for programmers and such, so it may be possible to get a job in that way, but I wouldn't bet on it.

    If you want to change to physics, why are you doing a computing PhD?
  6. Aug 7, 2010 #5
    I have been quiet fascinated with graph theory, hence I am doing a Phd in Graph Theory. But at the same time I am also interested in Physics but I must confess it is a relatively new interest which has developed. Being a person from a non-physics background, I sought to elicit the answers from this forum. Which, I must say, has been very informative (thanks to you)
  7. Aug 8, 2010 #6
    The only thing that I can think of off-hand is optimizing N-body simulations on parallel computers.

    Possible. Not easy. Probably the easiest thing to do is to get a job at some supercomputing center where there is general research on numerical applications.

    I think there are easier ways of getting what you want.
  8. Aug 8, 2010 #7
    I don't think so. What you do what to do is to network and talk with astrophysicists that might be interested in your research. Someone that occurs to me is that if you could take the output of the Los Alamos server and use graph theory to figure out what physicist that you should be talking to.

    You might want to take a few courses, but it's better to do that ala-carte rather than as part of a degree program.
  9. Aug 8, 2010 #8
    Something else that might be useful is to use graph theory to calculate Feynman diagrams for ultra high expansions. All of the QFT Feyman diagrams I've seen have been hand calculated, but it would be interesting to have some sort of math engine calculate the 200th order Feynman diagram expansion of something......
  10. Aug 9, 2010 #9
    Thank you for replying. I was not aware that one could take courses al-a-carte at a Uni :smile: and what did you mean when you said get the output from the LANL website ?
  11. Nov 18, 2011 #10
    Hi everybody,
    I have passed my M.Sc. in computer science. i want to go for a ph.d. Graph theory is a topic i have always been fascinated with. Can someone tell me where in europe can i do a Ph.D in the same with a scholarship ?
    Also it will be helpful if someone gives me information on ph.d in other computer science subjects, like, algorithm and data structures, programming, operating systems etc..
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook