|
PureBytes Links
Trading Reference Links
|
> -----Message d'origine-----
> De : Treasure [mailto:Treasure@xxxxxxx]
> Envoyé : dimanche 15 décembre 2002 07:33
> À : Gene Pope; omega-list@xxxxxxxxxx
> Objet : RE: Limited life span of mechanical systems?
>
>
> Right, only what we are searching for, is the Neural Net to really become
> predictive based on historical information. And really, isn't that what any
> system must do, analyses the past to predict the future. Otherwise the whole
> thing is a crap shoot. The NN does this (tries to be predictive based on
> past results). What seems to fail others who have tried NNs is that they are
> not breaking up the data into various configuration and point of views. The
> NN needs to see data from a variety of intelligently built and carefully
> designed data presentations.
>
> Treasure
>
This is where fuzzy logic has a serious advantage over NN.
The data prsentation , through fuzzy logic, is close to a case per case study,
where the rules are different according to the indicators ( inputs) values.
Thsi could be seen, when using neurofuzzy logic, as a NN applied in a different
manner according to the fuzzy logic membership defined by the indicator values
themselves. In fact, it's more complicated , but you got the general idea.
A more clever desing than the classical NN approach where the same NN has to
cope with the various indicators configuration.
Predicting or not is not an issue. We consider that the neurofuzzy logic rules
acts as a smart data filtering tool, with the minimum lag possible and the best
signal to noise ratio. What is enough to build winning trading systems, without
worrying about prediction issues.
Sincerely,
Pierre Orphelin
www.sirtrade.com
Tradestation 2000i, TradeStation 6 sales and support
Safir-X, neurofuzzy logic trading system builder
==>Special offer ends 12/24/2002.
To subscribe to our FREE trading system newsletter:
http://www.sirtrade.com/newsletter.htm
|