|
PureBytes Links
Trading Reference Links
|
-----Message d'origine-----
De : Lawrence Chan <stnahc@xxxxxxxx>
À : omega-list@xxxxxxxxxx <omega-list@xxxxxxxxxx>
Date : mardi 19 décembre 2000 01:27
Objet : Re: The Yats Group
>
>I had a not so nice arguement with PO about 9 months
>ago on this very issue.
>
I suspect that your thinking is biased when I read " Laurence Chan and PO" in the same message.
>simply put - if neural net can accurately predict
>the future all the time, then it is technically
>proving the concept of "P=NP" (in layman's term
>you can crack any high security code in no time
>and you can tell with 100% accuracy about tomorrow's
>weather. :) )
We predict nothing at all.
The test was successful because their so called random data serie was as random as MB is when speaking about Omega.
At iit's not NN.
It's fuzzy logic + NN ( main tasks being done by mean of FL)
>
>thus, if neural net is good at something, it will
>be at best par to any other tools available and
>the result will at most be "quite" accurate most
>of the time and definitely screw up at the worst
>time :)
Glups...
>
>it seems to me most people use NN the wrong way
>assuming the NN will do everything for them ...
>just thinking that the time it will take to train
>and retrain NN is about the same as doing other
>types of analysis anyway, I prefer doing the other
>ways :)
>
We are not most people, probably
>
>=====
>Lawrence Chan http://www.tickquest.com
>Innovative Analytical Software for Trading Professionals
What is this ?
Advertising on Omega List ?I cannot believe it
Rgds
PO
www.sirtrade.com
>
>
|