• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Question for programmers

X

xhail

Limp Gawd
Joined
Jun 21, 2002
Messages
187
I'm a college student currently enrolled in CS courses for a CS degree. I was just curious as to the extent that Numerical Analysis is used in real world jobs. For my school, this class entails approximations of functions, differentiation, integration, interpolation and systems of equations. The main highlight being the error in each of these topics due to approximating a function in order to do one of the methods.

Could someone please tell me how often something like this is used?
 
Depends on what you do. "CS" as a field is much too generic. I'm working on flight navigation software for helicopters and use spline interpolation routinely for navigation plus a little extrapolation for threat regions.

Nearly anything involving real-world modelling will rely on numerical analysis, as the real world is too complex to model precisely.
 
Newer games (especially First Person Shooters) tend to involve TONS of mathmatics in general, esspecially when the add physics into the mix.
 
Let me see if I can clarify my question a bit.

Is there a need to know all the proofs and theories that these methods are based on? Or is it more important to know how, when and where to do these type of methods?

In terms of CS, I want to work/research A.I. (I have a lot of schooling left). When I have time, I also want to make small game engines just as a hobby/fun. What kind of mathematical methods are used for these two fields?

Thanks
 
research == need to know proofs. Otherwise it's sorta debatable.
 
If I recall correctly most of the AI stuff I've looked at involves a lot of set theory, searching, and pattern matching. Probably some sorting too. Trigonometry and physics are going to be important in a game, at least if it's 3D.
Focus on CS theory while you're in school. Once you get to the "real world" you'll find that they probably only taught you a little about actually programming. What you'd call a big project in undergrad is a little project in the real world, and a big real world project would take you all four years of college to write. Plus the languages and such that you pick up in school will go out of style eventually. Theory just gets added to for the most part.
The catch is you'll never be a really good programmer if you don't understand the theory behind it. You'd be amazed at how many candidates I've shot down for an open position because they didn't know the difference between a tree and a hash table.
As for numerical analysis, it really depends on your field. For what I'm doing now, basically zip. Of course, I'm working in an eCommerce job. At my last job I wrote some scientific imaging software. There I did a whole lot of number-crunching, and understanding numerical analysis is useful.

edit: BTW I'm a professional software engineer.
 
xhail said:
What kind of mathematical methods are used for these two fields?
Click to expand...

What's the point in writing new engines if you're just going to do the same thing everyone else does? You've gotta take the classes anyways, learn the shit and someday you might find a new way in which it's applicable to the problem domain.
 
For my part, I'm a maths graduate turned programmer by trade, and I can tell you honestly that I'm glad I did it that way round. Number theory may not be particularly interesting in itself, but it'll cure you of the fear of "difficult number stuff" that a lot of programmers have. That kind of thing is far harder to learn on your own than a new programming language. While I don't use much of the specific stuff I learned in university (fluid mechanics, system theory etc), it gave me the ability to just look at a problem as a problem, not a completely alien set of circumstances where I don't know where to begin.

In short, lap it up, kid, because it'll make your life much easier in time to come ;)
 
To expand on what :LJ: said... I know of several math graduates who ended up teaching or studying computer science. I don't know many who went the other way 'round.

At my university, you only need a couple of extra classes outside the CS curriculum to get a math minor or major. For me, it was worth it to take the extra classes, since one of them--numerical methods, of which you're inquiring--was a double-dip, counting as electives in both math and computer science.
 
I'm going through all my math courses including this one and doing well. I am trying to learn it instead of "learn it for the exam". It's just that the professor does not know how to teach. I usually rely on the text book and any other sources of information I can get my hands on. Thinking of plunking down some cash for a good book dealing with said subject, but couldn't really find one that encompassed all the different methods discussed in class. Instead, I would probably have to invest in a couple math textbooks. Money being tight as it is, I wanted to get a good feel for the importance of the material before plunking down hard earned cash :)

Thanks for clearing things up. Guess I'll go and buy those books!
 
Be sure to look over those books before you drop the cash. A lot of math is useful in programming, a lot isn't. I've never had any use for differential geometry in programming, but numerical methods are definately useful and chaos theory (fractal compression algorithms in particular) can be as well. Now granted they're not going to teach a fractal compression algorithm in a typical chaos theory class, but if you don't understand the theory good luck figuring out how the algorithm works. If you have the time and inclination to study this stuff more in depth than the class requires by all means go for it. Some day it'll come in helpful, and if you're really good at math and programming perhaps you can land a job in the finance industry (big $$$$). But if the extra studying will hurt you in another class I'd save your cash for now. If you can learn it from a book now you can learn it from a book later.
 
I'm sure you would need to know that stuff if you worked for company's that make design and simulation programs.. you know stuff like MatLab, Math Cad, PSPICE. etc... crap loads of other programs...

Matlab is good for vision processing (that's machine vision).. so it's defiantly related to AI i would say.. but i'm not a program.. i'm on the hardware side of the spectrum..
 
You must log in or register to reply here.
Back
Top