Machine Learning is the new C++

Discussion in 'App Development' started by xandman, May 28, 2015.

  1. I think there has to be some user input to tell the AI what the meaning of 'good' is. So you can teach an AI to play chess by letting it play a large number of repeated games against a human or chess computer, and telling it ONLY (A) when it is moving illegally, and (B) when it has won or lost.

    You can give it more information such as 'this position looks good' or 'a knight is worth about the same as a bishop'. This will speed up the learning process, but its not strictly necessary, and by doing so you're putting your own biases and ideas into the process.

    Similarly as a minimum I'd expect to have to tell the AI that (a) I want to maximise risk adjusted return and (B) how I define risk. This isn't unreasonable as there is no way the AI would know what my utility function is (as defined in B) - even another human wouldn't know that. But you shouldn't have to put any more than this in. To continue the analogy you shouldn't have to tell it 'stocks are usually uncorrelated with bonds'. That is both forward information, and a bias.

    So I agree with you that entering a fine tuned utility function that will recover the users prejudice, or put implicit forward information in, is definitely not a good idea.
     
    #21     Jun 2, 2015
    dartmus, IAS_LLC and eusdaiki like this.
  2. Banjo

    Banjo

    Context is ever changing in nano degrees re: markets. It seems impossible to capture this in order to define/refine rule sets, especially in an ongoing manner.
     
    #22     Jun 2, 2015
  3. IAS_LLC

    IAS_LLC

    Yup, we are saying the same thing. I'm a big fan of AI/ML algorithms.. but without me, they are worthless. Me + AI results in capability that is greater than the sum of its parts.
     
    #23     Jun 2, 2015
    dartmus likes this.
  4. Banjo

    Banjo

    #24     Jun 4, 2015
    xandman likes this.
  5. Sergio77

    Sergio77

    I knew this guy who ran a fund and he thought the same. Now he is Uber driver:)
     
    #25     Jun 4, 2015
  6. IAS_LLC

    IAS_LLC

    Please explain
     
    #26     Jun 4, 2015
    dartmus likes this.
  7. Sergio77

    Sergio77

    #27     Jun 6, 2015
  8. IAS_LLC

    IAS_LLC

    #28     Jun 6, 2015
    dartmus likes this.
  9. Humpy

    Humpy

    Virtu systems are claiming a 95%+ win rate using HFT and computer trading.
    Could this be the end of the independent trader ?
     
    #29     Jun 6, 2015
  10. only if all traders trade in the same timeframe as Wheretoo
     
    #30     Jun 6, 2015