3

My interests in CS are fairly narrow; I'm mainly interested in numerical analysis and numerical methods for solving differential equations. However, during my undergrad I mainly focused on other, more traditional aspects of CS. I don't have as strong a preparation in the applied math side required for a PhD in Applied Math/Numerical Analysis. A class I took my senior year (second semester!) got me really interested in numerics and I've spent much of my free time since then self-studying the material.

Are there any good CS programs where I can get a good background in applied math/numerical analysis and largely avoid topics like graph theory and combinatorics? With my (relatively) weak preparation in formal math, I think a good math masters is not in the cards.

Emir
  • 2,213
  • Have you considered an advanced degree in "Computational Science"? – Amzoti Aug 25 '13 at 23:46
  • From my research it seems that very few schools offer such a specialized degree. – Emir Aug 26 '13 at 14:08
  • But if that is what you want to do, why not move there? – Amzoti Aug 26 '13 at 14:16
  • You have a good point, and I'm willing to make the move if need be. My eventual goal is to do a PhD in the field and I'm not sure how well these programs place. Many of them seem geared for industry-bound engineering types. – Emir Aug 26 '13 at 14:46
  • You should ask a trusted professor about this. They will have more information about your preparation and abilities, so they will be able to advise you much better than anonymous people on the internet ever could. – Potato Aug 28 '13 at 01:30
  • I am working on a posteriori error estimation in differential form finite elements for Maxwell's equations, I happen to know some interdisciplinary PhD programs under the computational science centers. Many of them are doing researches in how to solve large scale partial differential equation problems (electromagnetics, fluid dynamics, or even multiphysics simulation). Are you interested in this? – Shuhao Cao Aug 28 '13 at 01:30
  • Shuhao, sounds potentially interesting. Do you see that most computational science PhDs work on applied numerical analysis rather than theory? Is theoretical NA a mostly complete field? – Emir Aug 28 '13 at 18:45
  • Theoretical numerical analysis is sure a completely field, as well as the simulation/implementation part. Recent trends include but not limited to: designing robust model for multiphysics, large scale linear system solvers like LAPACK but more problem oriented (e.g. Maxwell, Navier-Stokes), parallel/high-performance/gpu computing, adaptive mesh generation/automated PDE solution, inverse problems. – Shuhao Cao Aug 30 '13 at 05:19

2 Answers2

1

Many schools offer masters in programs that are not specific to just Cs and math. If you're interest in CS is "fairly narrow" as you described it, I wouldn't look into a masters in CS, I would look into a masters in data computing/statistics or even financial Mathematics. Most of these topics do not require combinatorics and graph theory as pre-reqs or requirements.

waj cheema
  • 129
  • 1
  • 2
  • 9
0

There is a program offered by UCSD which is a specialization of Computer Science and Applied Mathematics. It is supposed to be a really good program.

RDizzl3
  • 3,207