Mechanical vs Neural trading systems

Discussion in 'Automated Trading' started by pandabear, Nov 11, 2007.

  1. It's simpler...

    Most people just stick a NN or Computer Learning logic and lose control over it.

    It's a necessity for a system developer to have clarity over what's going on with their trades.

    Blindly using NN or any tool without any understanding / control is only increasing your risk of trading.
     
    #11     Nov 14, 2007
  2. ronblack

    ronblack

    1. Yes

    2. "Neural system" is another term for curve fitting when it applies to market data. The biggest problem is overtraning.

    Ron
     
    #12     Nov 14, 2007
  3. lindq

    lindq

    LOL. Not the most powerful offer I've ever heard.

    Don't wait by the phone.
     
    #13     Nov 14, 2007
  4. lindq

    lindq

    LOL. Not the most powerful offer I've ever heard.

    Don't wait by the phone.
     
    #14     Nov 14, 2007

  5. Where should he wait?
     
    #15     Nov 14, 2007
  6. Thanks for a lot of these comments! Very helpful indeed.

    A couple of followup questions:

    - Most folks who are running profitable mechanical trading systems - Are you running intraday or swing systems? Are you running them on SPY, DIA, IWM's or looking on individual stocks?

    Thanks
     
    #16     Nov 25, 2007
  7. Sure, that is why all the volume keeps moving more and more to Automated Systems, because everyone is just losing in all these systems....LOL!!!!!!!!!! :D

    Remember, just because you can't build your own bridge across the river does not mean others can't.
     
    #17     Nov 25, 2007
  8. Swing.
    SPY, DIA, and individual stocks.

    Running Mechanical NN systems profitably.
     
    #18     Nov 25, 2007
  9. Just another comment.

    It would not surprise me if the Naysayers here on mechanical trading systems or NN systems or mechanical systems enhanced with NN's have never actually worked with them.
     
    #19     Nov 25, 2007
  10. nitro

    nitro

    Agreed.

    nitro
     
    #20     Nov 25, 2007