Yeah, I can see how those things would definitely help, and I didn't know that FIPS certification required a finite state machine. I don't deal with the hardware side, and really, the only algorithm I've studied in any great detail (read: tried to do limited cryptanalysis on) is Rijndael/AES. For my end though, graph theory is useful for understanding and modifying things like, Simple Temporal Networks, which CMU has applied towards resolving timing issues in multi-agent systems. I have to say though, I've had absolutely zero opportunity to use formal grammars since learning them 7 years ago.... given the way formal grammars actually work though, I'm not sure if I will ever explicitly use the information that I learned there, or if taking that course was enough to simply influence the way that I view/use computers. So I would concede that formal languages and automata constitutes a good choice regardless of whether or not you'll ever need to use what you learned as a "tool" since the subject kinda goes a bit deeper than that.
Statistics is something that seems to be under-emphasized for CS students though. I never really applied myself in those courses and I kinda regret it now. It seems like every time I turn around I'm looking at a problem that has a slick solution if I view the problem as a stochastic process, solve that using statistical tools and then bring the result back into the "real world". Having taken the classes, I've got a fairly solid grounding, so I'm not too bad off... but I wish I'd put more effort into those ones.
I have to contest one point though, the whole "cornerstone of good coding" thing. It's probably a misunderstanding (on your part, caused by me failing to actually write what I meant....) but I stick with my statement about linear algebra. In most cases, anything written in lin-alg-speak translates extremely well to pretty much any language (cept perhaps befunge and the like...) CS is about coming up with novel ways of solving problems (the science) and then implementing/validating those ideas using a computer... to me, it follows quite nicely that if your ideas are formulated using linear algebra, you're going to have that much of an easier time validating them using code.... You may not entirely agree because of what you work on, but since most of the crypto algorithms that you'd be building off of are written using linear algebraic techniques, you might... Coding is an important part of CS though. A computer scientist is a mathematician who specializes in applying math using computers. I mean, just look at your field, crypto is a mathematical endeavour... most of the algebraicists I know are very much into crypto since it's all written in their language. It's a series of mathematical operations applied to a matrix... it's designed so that you could do the operations by hand (except RSA;-), but then it's optimized to run on a computer.... to me, crypto is CS at its finest... but at all levels and not just the hardware/software implementation level.
I'd have to agree with the beginning of your post: Calculus isn't all that useful to a CS major. Don't get me wrong, calculus is useful in and of itself, but since computers can't do integrals or derivatives, it's kinda useless. Calculus combined with something like numerical methods, which teach you how to use various methods of approximation and interpolation to get a computer to understand calculus...well... that's a different story. Even then though, I don't tend to use calc or numerical methods for much. Algebra and Galois Fields tend to be the maths I use the most. GFs are especially nice for crypto since GF(2^8) and the operations that are closed over it represent everything you can do to a byte...
Mind you, I'm mostly interested in ai and crypto. Crypto is mostly a hobby, but for ai having a rock solid understanding of lin alg helps in performing things like computer vision (lin alg for the pixel manipulations, trig ftw for the angular calculations), path planning (okay, some calc, but mostly lin alg as we're, again, approximating a continuous curve over a discrete space, so it's summations instead of integrals) motion prediction (same as last)... etc...
So again, I'd have to say, Algebra is the math that a CS major should be the strongest in. The formal languages, automata, graph theory, etc... is all fairly standard CS stuff. None of it is as much a cornerstone of good coding as lin alg IMHO, but they're all really really nice to have sometimes.
As for the rest of your post, it's pretty obvious that your experience has been with parsers and compilers (or something similar), in which case, the rest of your post is 100% true. However, for my corner of the CS world (ai and robotics) lin alg and trig are t3h sh!t... followed by probs/stats (for stochastic modelling/simulations)....
I guess a "YMMV" belongs at the end of each post in this topic as everyone needs different math tools based on what they're working with.
This should provide a bit of an answer to both the OP and the people asking "what do you need maths for if you're going to be a sysadmin?" (You're either trolling or confused as to what CS guys actually do, so I'll give the benefit of the doubt to you and shed some more insight into what some of us do...)
Though many people would seem to be of the opinion that animal life is more important than human life... to them I would suggest that they put "their money where their mouth is" and feed themselves to the nearest pit-bull.
If you (not you eighty4) revere animal life as being so sacred, and consider human life to be as base as it gets, then perhaps you should attempt to remedy that situation by, as I said, feeding yourself to an animal... or is animal life only more important than human life when that human life is not your own?
These are not labs full of megalomaniacs remaniscent of the Joker from Batman, methodically torturing poor kittens and other small furry animals. They're performing scientific tests with as much humanity as they can. The fact that these tests, if performed upon a human, would still be cause for riots and lynching is an unfortunate by-product of the world we live in and the diseases these researchers are trying to cure.
If suddenly it were determined that rats were actually sentient on more than a simple stimulus-response with limited memory, then I'm sure you'd have a LARGE number of researchers who would repent and find another animal whose self-awareness is more in doubt. If you think that this is cruel and calculating, well... in a way... you're right. Welcome to the real world, it's not all rainbows and songbirds.
Exactly... if this were iron-clad and applied even to code that wasn't yours, there would be nothing to stop an MS employee from posting the source code to the entire and then using it as if it was not GPL'ed... which is clearly silly...
Ahh... y'know... that's even more annoying than the small conspiracy theory I put forth. I didn't know Mono was that widely shunned/avoided. It does seem odd that, as you say, a failing product geared towards the FOSS community is likely getting revamped through the support of a very-publicly-anti-FOSS organization. Thanks for the info... and who says you don't learn anything on/.?;-)
I haven't really been following Mono, or.NET, but upon reading this I was suddenly reminded of the story from a couple days ago about Novel and Microsoft getting into bed together.
Now, I'm sure a number of anti-microsoft fanboi-types will automatically jump all over this, but I'm hoping that someone who isn't a member of that group can explain to me if.Net (and sorta by extension, Mono) is a big enough deal to Microsoft that they would worm their way into this solely for the purposes of shitcanning Mono. Is it big enough? Is it possible? Would they even want to?
Lol... as politically incorrect as that is... it's actually kinda funny:-)
I know I'll get modded down for simply voicing my opinion on this, but I figured someone had to let you know that your snappy one-liner made someone chuckle.
A friend of mine had that same thing happen once, it started to infect other machines around it; we had to nuke the site from orbit... it was the only way to be sure...
Oh great, well, since you seem to have so much experience in the field, I have a problem that maybe you could help me with:
I'm trying to use a genetic programming technique like PIPE to generate parse trees capable of controlling a robot playing soccer. The problem is that the simulator I'm using is the RoboCup SoccerServer. You can't really speed it up since it's a real time simulation, and I really want to have the resultant algorithm be capable of winning in that environment, so I don't want to speed it up.
Typically, PIPE, with a generation size of 1000 genomes, requires somewhere around 5000 iterations to become any good at playing soccer. With the way PIPE, and most genetic programming algorithms in general, are structured this is unavoidable. However, each iteration is a complete 5 minute game, which means that I need 1000*5000*5 = 25 million minutes worth of simulation time (about 17000 days) to come up with, what should be, a GOOD parse tree capable of playing soccer.
Since you've been working in IT for the last 15 years or so, and since IT is obviously the exact same thing as comp sci, I was wondering if you could help me out by helping me find a way to make GP techniques converge faster?
Comp Sci is more math than anything else. I understand what the guys in IT do, cuz I've done system administration in both windows and linux (in 5000 user campus environments) as well as Network Admin in the same environments. My B.Sc. in Comp Sci have helped me tremendously in those environments, but I can guarantee you, now that I'm working on my M.Sc. in Comp Sci, that not a single one of those sysadmins and netadmins could understand even a tenth of the material that I'm required to learn....
Straight computer science? Well, you'll probably be just a code monkey.
I'm not sure where you guys all went to school, but it sure wasn't where I'm at.
All the comp sci guys that I went to school with (including meself eh) ended up being roughly equal parts code monkey and mathematician. Not as good at coding as a software engineer (though some were better...) and not as good at math as a pure math geek (though, again, some were better...)
Is my university that different, or are people just going by casual observation? I mean, I'm working on my M.Sc. in Comp Sci, and so far it's mostly math that I've been doing... I'm only just starting to build a simulation of what I sketched out mathematically so that I can get some data to analyze, but still, I'm at around 75% research and math, and 25% coding.
For anyone who cares, I'm trying to find out whether or not the premises behind the design choices of Probabilistic Incremental Program Evolution (PIPE, by Salustowicz and Schmidhuber) hold true in the RoboCup domain.
... usually a pedant... but you don't train a database. It was likely a neural net, but TFA is rather thin on details. Anyone got a link to their paper?
You have tons of pictures made by light waves travelling at the speed of light, but you haven't got a single picture of a light wave travelling at the speed of light.
Slashdot Mirror
.... tough crowd :-(
Yeah, I can see how those things would definitely help, and I didn't know that FIPS certification required a finite state machine. I don't deal with the hardware side, and really, the only algorithm I've studied in any great detail (read: tried to do limited cryptanalysis on) is Rijndael/AES. For my end though, graph theory is useful for understanding and modifying things like, Simple Temporal Networks, which CMU has applied towards resolving timing issues in multi-agent systems. I have to say though, I've had absolutely zero opportunity to use formal grammars since learning them 7 years ago.... given the way formal grammars actually work though, I'm not sure if I will ever explicitly use the information that I learned there, or if taking that course was enough to simply influence the way that I view/use computers. So I would concede that formal languages and automata constitutes a good choice regardless of whether or not you'll ever need to use what you learned as a "tool" since the subject kinda goes a bit deeper than that.
Statistics is something that seems to be under-emphasized for CS students though. I never really applied myself in those courses and I kinda regret it now. It seems like every time I turn around I'm looking at a problem that has a slick solution if I view the problem as a stochastic process, solve that using statistical tools and then bring the result back into the "real world". Having taken the classes, I've got a fairly solid grounding, so I'm not too bad off... but I wish I'd put more effort into those ones.
I have to contest one point though, the whole "cornerstone of good coding" thing. It's probably a misunderstanding (on your part, caused by me failing to actually write what I meant....) but I stick with my statement about linear algebra. In most cases, anything written in lin-alg-speak translates extremely well to pretty much any language (cept perhaps befunge and the like...) CS is about coming up with novel ways of solving problems (the science) and then implementing/validating those ideas using a computer... to me, it follows quite nicely that if your ideas are formulated using linear algebra, you're going to have that much of an easier time validating them using code.... You may not entirely agree because of what you work on, but since most of the crypto algorithms that you'd be building off of are written using linear algebraic techniques, you might... Coding is an important part of CS though. A computer scientist is a mathematician who specializes in applying math using computers. I mean, just look at your field, crypto is a mathematical endeavour... most of the algebraicists I know are very much into crypto since it's all written in their language. It's a series of mathematical operations applied to a matrix... it's designed so that you could do the operations by hand (except RSA ;-), but then it's optimized to run on a computer.... to me, crypto is CS at its finest... but at all levels and not just the hardware/software implementation level.
Oh Random Slashdot Quote at the bottom of the page... sometimes you are just soooo apropos:
hard, adj.: The quality of your own...etc... ;-)
I'd have to agree with the beginning of your post: Calculus isn't all that useful to a CS major. Don't get me wrong, calculus is useful in and of itself, but since computers can't do integrals or derivatives, it's kinda useless. Calculus combined with something like numerical methods, which teach you how to use various methods of approximation and interpolation to get a computer to understand calculus...well... that's a different story. Even then though, I don't tend to use calc or numerical methods for much. Algebra and Galois Fields tend to be the maths I use the most. GFs are especially nice for crypto since GF(2^8) and the operations that are closed over it represent everything you can do to a byte...
Mind you, I'm mostly interested in ai and crypto. Crypto is mostly a hobby, but for ai having a rock solid understanding of lin alg helps in performing things like computer vision (lin alg for the pixel manipulations, trig ftw for the angular calculations), path planning (okay, some calc, but mostly lin alg as we're, again, approximating a continuous curve over a discrete space, so it's summations instead of integrals) motion prediction (same as last) ... etc...
So again, I'd have to say, Algebra is the math that a CS major should be the strongest in. The formal languages, automata, graph theory, etc... is all fairly standard CS stuff. None of it is as much a cornerstone of good coding as lin alg IMHO, but they're all really really nice to have sometimes.
As for the rest of your post, it's pretty obvious that your experience has been with parsers and compilers (or something similar), in which case, the rest of your post is 100% true. However, for my corner of the CS world (ai and robotics) lin alg and trig are t3h sh!t... followed by probs/stats (for stochastic modelling/simulations)....
I guess a "YMMV" belongs at the end of each post in this topic as everyone needs different math tools based on what they're working with.
This should provide a bit of an answer to both the OP and the people asking "what do you need maths for if you're going to be a sysadmin?" (You're either trolling or confused as to what CS guys actually do, so I'll give the benefit of the doubt to you and shed some more insight into what some of us do...)
DAMN YOU!!!! Damn you and your prisms!!!111oneoneone *shakes fist*
Though many people would seem to be of the opinion that animal life is more important than human life... to them I would suggest that they put "their money where their mouth is" and feed themselves to the nearest pit-bull.
If you (not you eighty4) revere animal life as being so sacred, and consider human life to be as base as it gets, then perhaps you should attempt to remedy that situation by, as I said, feeding yourself to an animal... or is animal life only more important than human life when that human life is not your own?
These are not labs full of megalomaniacs remaniscent of the Joker from Batman, methodically torturing poor kittens and other small furry animals. They're performing scientific tests with as much humanity as they can. The fact that these tests, if performed upon a human, would still be cause for riots and lynching is an unfortunate by-product of the world we live in and the diseases these researchers are trying to cure.
If suddenly it were determined that rats were actually sentient on more than a simple stimulus-response with limited memory, then I'm sure you'd have a LARGE number of researchers who would repent and find another animal whose self-awareness is more in doubt. If you think that this is cruel and calculating, well... in a way... you're right. Welcome to the real world, it's not all rainbows and songbirds.
Exactly... if this were iron-clad and applied even to code that wasn't yours, there would be nothing to stop an MS employee from posting the source code to the entire and then using it as if it was not GPL'ed... which is clearly silly...
Ahh... y'know... that's even more annoying than the small conspiracy theory I put forth. I didn't know Mono was that widely shunned/avoided. It does seem odd that, as you say, a failing product geared towards the FOSS community is likely getting revamped through the support of a very-publicly-anti-FOSS organization. Thanks for the info... and who says you don't learn anything on /.? ;-)
I haven't really been following Mono, or .NET, but upon reading this I was suddenly reminded of the story from a couple days ago about Novel and Microsoft getting into bed together.
Now, I'm sure a number of anti-microsoft fanboi-types will automatically jump all over this, but I'm hoping that someone who isn't a member of that group can explain to me if .Net (and sorta by extension, Mono) is a big enough deal to Microsoft that they would worm their way into this solely for the purposes of shitcanning Mono. Is it big enough? Is it possible? Would they even want to?
Get the State out of marrige. (Score:5, Oh Snap!)
Given the state of our educational system, they'll probably be more qualified anyway.
New mod option anyone? ;-)
Lol... as politically incorrect as that is... it's actually kinda funny :-)
I know I'll get modded down for simply voicing my opinion on this, but I figured someone had to let you know that your snappy one-liner made someone chuckle.
Didn't I see this story in a Jenna Jameson flick once?
A friend of mine had that same thing happen once, it started to infect other machines around it; we had to nuke the site from orbit... it was the only way to be sure...
you're right, they should just not try instead ;-)
some ideas pan out... some don't. Hell, these little things might even become horribly useful with some minor tweaking.... . . . .. and so it marches.
/grin ... indeed
Oh great, well, since you seem to have so much experience in the field, I have a problem that maybe you could help me with:
I'm trying to use a genetic programming technique like PIPE to generate parse trees capable of controlling a robot playing soccer. The problem is that the simulator I'm using is the RoboCup SoccerServer. You can't really speed it up since it's a real time simulation, and I really want to have the resultant algorithm be capable of winning in that environment, so I don't want to speed it up.
Typically, PIPE, with a generation size of 1000 genomes, requires somewhere around 5000 iterations to become any good at playing soccer. With the way PIPE, and most genetic programming algorithms in general, are structured this is unavoidable. However, each iteration is a complete 5 minute game, which means that I need 1000*5000*5 = 25 million minutes worth of simulation time (about 17000 days) to come up with, what should be, a GOOD parse tree capable of playing soccer.
Since you've been working in IT for the last 15 years or so, and since IT is obviously the exact same thing as comp sci, I was wondering if you could help me out by helping me find a way to make GP techniques converge faster?
Thanks a mil,
MOD PARENT UP!!!
Comp Sci is more math than anything else. I understand what the guys in IT do, cuz I've done system administration in both windows and linux (in 5000 user campus environments) as well as Network Admin in the same environments. My B.Sc. in Comp Sci have helped me tremendously in those environments, but I can guarantee you, now that I'm working on my M.Sc. in Comp Sci, that not a single one of those sysadmins and netadmins could understand even a tenth of the material that I'm required to learn....
Again, COMP SCI != IT
... finally a story that deserves the "itsatrap" tag
I'm giving 10 to 1 odds for the next 5 minutes, that it won't get it ;-)
Straight computer science? Well, you'll probably be just a code monkey.
I'm not sure where you guys all went to school, but it sure wasn't where I'm at.
All the comp sci guys that I went to school with (including meself eh) ended up being roughly equal parts code monkey and mathematician. Not as good at coding as a software engineer (though some were better...) and not as good at math as a pure math geek (though, again, some were better...)
Is my university that different, or are people just going by casual observation? I mean, I'm working on my M.Sc. in Comp Sci, and so far it's mostly math that I've been doing... I'm only just starting to build a simulation of what I sketched out mathematically so that I can get some data to analyze, but still, I'm at around 75% research and math, and 25% coding.
For anyone who cares, I'm trying to find out whether or not the premises behind the design choices of Probabilistic Incremental Program Evolution (PIPE, by Salustowicz and Schmidhuber) hold true in the RoboCup domain.
Netcraft confirms it?
... usually a pedant... but you don't train a database. It was likely a neural net, but TFA is rather thin on details. Anyone got a link to their paper?
hahaYouhahaFail.
Did I not RTFA correctly, or is Satan breaking out his ice-skates??
Does this mean that SUSE is now as good as dead, or poised to take over?
I don't know if this would enable you to break a hole into a hell dimension, but it certainly sounds plausible,
Uh oh.... I smell a DOOM sequel...
As has been pointed out already, no, you don't.
You have tons of pictures made by light waves travelling at the speed of light, but you haven't got a single picture of a light wave travelling at the speed of light.
there's a difference...