Wavelet decomposition ... and noise

H

harrytrader


The picture below was in February 2001 issue of WIRED magazine according to this article http://www.bearcave.com/misl/misl_tech/wavelets/freq/
<IMG SRC=http://www.elitetrader.com/vb/attachment.php?s=&postid=431481>
I can tell you from the point of view of Walter Shewart's Statistical Process Control this cannot be considered as noise in the sense noise = many independant little causes. This is why today's some technoloclasts because they have computers and mathematical tools in their hand think the world is reduced to their model and so consider that the rest is noise whereas it is their models that are just false haha ! It is not the fault of wavelet decomposition itself (wavelet is an evolution of cycle analysis techniques of the same kind than fourier but with varying windows lenght) it is the fault of how it is inadequately used.

This is why understanding what randomness is can avoid such error that's why I had asked the question:
"Do you really know what Randomness is ?"
http://www.elitetrader.com/vb/showthread.php?s=&threadid=26294&highlight=noise
 

Attachments

Thisi s the best elementary description I have ever seen of this. Thanks for the link.

I will comment on the connection to randomness later.

nitro
 
Quote from harrytrader:

The picture below was in February 2001 issue of WIRED magazine according to this article http://www.bearcave.com/misl/misl_tech/wavelets/freq/
<IMG SRC=http://www.elitetrader.com/vb/attachment.php?s=&postid=431481>
I can tell you from the point of view of Walter Shewart's Statistical Process Control this cannot be considered as noise in the sense noise = many independant little causes. This is why today's some technoloclasts because they have computers and mathematical tools in their hand think the world is reduced to their model and so consider that the rest is noise whereas it is their models that are just false haha ! It is not the fault of wavelet decomposition itself (wavelet is an evolution of cycle analysis techniques of the same kind than fourier but with varying windows lenght) it is the fault of how it is inadequately used.

This is why understanding what randomness is can avoid such error that's why I had asked the question:
"Do you really know what Randomness is ?"
http://www.elitetrader.com/vb/showthread.php?s=&threadid=26294&highlight=noise
More...

=====

Harrytrader;

Like those multi year mountain charts with the 50dma;
just looked at some more mountain charts with the more conventional DIA and $INDU with bid/ask price at the hard right edge.

If you had asked me the elementary question would you prefer to study noise or more ''denoising''???;

prefer to study more '' denoising'':cool:

=====

More in 2004:cool:
 
Quote from nitro:

Thisi s the best elementary description I have ever seen of this. Thanks for the link.

I will comment on the connection to randomness later.

nitro
More...

Hey nitro,

When you've done your first trade with this, kindly send us a postcard. :D

Be good,

nononsense
 
...More stating of the obvious regarding "randomness".

Wavelets are just another technique - and not much better than other techniques except in special cases/applications.
 
What is the difference between wavelets, fourier, spectrum analysis and just neural nets?

Are not neural nets aimed to search for self similarity as patterns or cycles?

Of course the process can take a very long time for a NN with this purpose.
 
Thanks.

Now at the very beginning of my researchs years ago I have lost much time with these kind of "sophisticated" tools like neural net and frequency analysis: what counts is the model not the tools. Maybe now that I had a true model would I use again these tools upon my model but without a model it's worthless - as I said it would like a monkey playing with excel functions. Nevertheless I have already much stuffs to cope with at the moment so it would be for later.

BTW I found something funny on one of the site referenced :D

Who Owns Nature's Patterns?
An Analysis in Hindsight...
From "News of the Weird."

http://www.amara.com/stranger/str_fiction.html

In April 1997, Sir Roger Penrose, a British math professor who has worked with Stephen Hawking on such topics as relativity, black holes, and whether time has a beginning, filed a copyright- infringement lawsuit against the Kimberly-Clark Corporation, which Penrose said copied a pattern he created (a pattern demonstrating that "a nonrepeating pattern could exist in nature") for its Kleenex quilted toilet paper. Penrose said he doesn't like litigation but, "When it comes to the population of Great Britain being invited by a multinational to wipe their bottoms on what appears to be the work of a Knight of the Realm, then a last stand must be taken."

and less funny :(

You Think You Have Budget Problems...
Ottawa, Canada:

The keepers of a $70-million atom-shooting device in Chalk River have an explosive situation on their hands.

There's not enough cash to keep the machine running. Shutting it down will destroy it. And trying to keep the equipment running after skilled staff are laid off next month could cause it to explode.

Division director John Hardy says he's reluctant to pull the plug. If the cyclotron warms up from its deep freeze of -270 C, the temperature change will make its insides crack and leak, rendering it a $70-million junk pile.

But without experienced workers, the 1,000 gallons of liquid nitrogen used to cool the device may heat up and explode.


Quote from bdixon619:

for all you wavelet fans. This is heavier lifting than I care to do. However, I'm always interested in reading about these kinds of processes.

http://www.foretrade.com/Wavelet.htm
More...
 
You must log in or register to reply here.
Back
Top