Information Theory and Music.

Isn't that just how a regular person would look at it, who knows a few little things about thoery... like me? haha. I say yes!

Sorry guys it's not my intention of starting a flame war but I smell a rat. It wasn't my intention of stamping out this thread either. In fact I support an algorithmic approach to some degree(and I'm writing software that will eventually have such algorithms). The problem is, in useful algorithms will be quite complex and will have many modes of failure. My point is that DOFTS doesn't realize how complex it is... or maybe he's just a genius like he claims.

But to point out I didn't start the flame!

In any case I was going to mail this to him privately but decided to do it publicly. The point being that I cannot prove he has such a resume but I can definitely get some idea if he's a fraud. Not that I care too much but he started it!! ;) and he can finish it by answering the extremely simple questions below.

----------------------

With your resume I'd imagine you'd be more intelligent. To bad you can't back it up!

In fact! you can! You have a phd in applied mathematics?!? Here's 3 relatively simple problems and for someone like you should take no more than a minute to solve/prove.

1. sum(1/(k+9)^2) = ? (extremely simple weed out question... So simple even CAS can do it!)

2. int(f(x)*2^(i*w*t),t=-oo..oo) = ? (Another weed out question and it's not a trick question but simplified for your viewing pleasure)

3. Int(x^Pi*(1 - x)^(e-1) = ? (another weed out question but assumes you've actually had more than a year or two studying math)

Oh, you also studied physics?

1. What is e_j,k,q*E^i,j,k? (again, very simple if you have any real experience in physics.)

I guess I'll leave it at that. I do expect a quick and prompt reply with the solutions. Each problem is intentionally designed to determine if you have any experience in that area. I'll know from your response if you have a clue or not. (I intentionally avoided questions that could easily be looked up on google)

---------------

Seeing that he responds within a day I'll expect the answers by tomorrow. If he answers these positively then maybe we can actually get on with discussing potential algorithmic approaches to analysis and get something done instead of all the fluff. (Although I'm by far an expert in algorithms I have done a little bit imaging for a friend when he needed help and of course the routine stuff in CS)

But hard to talk about without really knowing anything about it. Any links to these analyses or studies?

I would still be interested in reading about whatever it is that we're talking about. Maybe it's just me. The proverbial special olympics are fun too, I guess.

But to point out I didn't start the flame!

But you're an idiot for insisting.

So please, if you don't agree, fine, but cut this absurd bullshit. I don't really give a hotdog, nor does anyone else, if you believe or not what DOFTS said about his penis so don't drag this into some absurd "HAHA SOLVE THIS MATHS!!!!" idiocy because it derails the damn thread.

So quit that crap, thx.

PS: Thought: Since we're applying crap to music analysis, I wonder if there's a place for game theory there somewhere...

*sigh*

So, basically you pick random equations without defining variables and in the first sum without giving bounds and tell me to solve them, cool.

Somehow you believe that the ability to do computational work is what makes me a mathematician?

You know want makes me a mathematician? A digraph D has a topological order if and only if D is acyclic.

Proof: If there is a topological order 1,2,...,n, then D could not have a cycle. Or Conversely, we suppose that D is acyclic. So Since every acyclic digraph has a vertex with no incoming arc there is a vertex with x1 with no incoming arc. Labeled x1 with the label 1, and delete it from D. Now the resulting digraph is still acyclic and hence has a vertex x2 with no incoming arc. We label x2 with label 2 and etc. Clearly given rise to a topological order.

3) o o o o o o...is that that thingy..um.....umm..what is it called? AHHHH I know it...AHHH transformation... LAPLACE! Is that it? It looks like it.

Your an idiot DOFTS. You can copy anything from google and that proves your a genius? I don't think so.

Don't define equations and terminology? If you had any clue and spent even 2 years seriously studying mathematics you would have the background.

You didn't even have to give exact answers but state you had some clue what was going on... which you obviously didn't.

1. Zeta function. Zeta(2) = sum(1/k^2). A simple shift in the index would give the answer.

2. Since f(x) is constant in the integration you have f(x)*int(e^(i*w*t)) which is the fourier transform of the dirac delta function. Even if f(x) was dependent on t you could have just wrote the fourier transform of f(t).

3. This one might be a bit harder but I'm sure with your 80 phd's you'd know that is just the beta function and is equal too gamma(pi)*gamma(e)/gamma(pi + e)

1. Those are just the permutation tensors. Surely you know what a tensor is? It's solution is just e_j,k,q or E^i,j,k since they are exactly the same!

---

In any case maybe you could use the excuse that the notation stumped you on one or two but not all.

Since you didn't even attempt to explain any of them, even ever so slightly, it's obviously you don't have much of a mathematical background. Maybe you shouldn't have exaggerated so much and it would have been more believable.

Hell, even Jubilee saw something and he/she doesn't have 183 phd's. Oh well.. I'll leave you people to discuss the complexities of algorithmic musical analysis. Have fun!

I'm his wife.

I'll leave you people to discuss the complexities of algorithmic musical analysis. Have fun!

K, if that means you'll stop posting gibberish, I'm totally for that.

Take your goddamn e-penis elsewhere, or contribute to the damn thread rather than lashing out because someone may actually have a better(?) academic background than you! (GASP!) Or whatever your problem is, nobody cares.

Jon, if you ever get to urge to bash me, just go to my journal, that way you can bash me, and you won't derail a thread at the same time, but anyways:

I've said this before, the algorithm isn't designed so that it looks for something exactly, it just looks.

Takes this example from breast cancer research. You can use segmentation to look for imagines that stick out. At first the image is broken up into little boxes and then it searches for boxes that are similar and then it groups all similar boxes to be one object. It would take a trained professional to tell you, "that's tumor." but nevertheless it's rather useful.

As to HOW it exactly does that, I have no idea how, information theory and this kind of program tends to be a field I don't fully understand, but I trust the input of better and more educated people than I.

As a side note:

Number 1 can be seen as a rather large fraction + a ploygamma function. Context matters, context is important.

I can't count how many times a DOFTS post starts with "Wow, that was the dumbest thing I have ever read, you truly are an idiot"

:happy:

What can I say, people keep on getting dumber!

DOFTS - please do not give up. It seems that Jon is overreacting to a TOOL. What I would love to hear from Jon is his knowledge of information and segmentation and what he knows about its applicability to analysis of music.

You both seem knowledgeable about areas of mathematics, it is a shame you both cannot cooperate a little more. Except for the original post from DOFTS, I haven't gained much from this as a non-mathematician -- it is a shame the subject fascinates me too bad everyone's values get in the way.

Your an idiot DOFTS.

...

3. This one might be a bit harder but I'm sure with your 80 phd's you'd know that is just the beta function and is equal too gamma(pi)*gamma(e)/gamma(pi + e)

...

Hell, even Jubilee saw something and he/she doesn't have 183 phd's. Oh well..

Wow, DOFTS you are smarter than I first thought. In the time It took Jon to write a dozen or so lines, you earned 103 phd's....

What can I say, I'm just motivated.

103 PhD's?

Umm...can I borrow one? A decent job sounds great to me

/me tosses TG a PhD in Economics. That should hold you for a while.