Machine learning services

Discussion in 'App Development' started by mag2717, Feb 5, 2024.

  1. mag2717

    mag2717


    Very very promising. Hopefully even amazing if done everything properly.
     
    #11     Feb 7, 2024
  2. Quanto

    Quanto

    IMO just marketing BS, nothing more :)

    Btw FYI: normally such commercial advertisements here at ET and in ET forums require a Vendor Account (aka Sponsor Account).
     
    Last edited: Feb 7, 2024
    #12     Feb 7, 2024
    murray t turtle likes this.
  3. Meh. Deep learning is very good at performing cognitive tasks that humans excel at, like recognizing cats in pictures. While there are similar pattern-recognition tasks in the markets (mostly in the HFT space), most of the longer term forecasting is based on tabular data and deterministic models are better suited for it. Also, feedback loops and salting are a real thing in finance, which makes RL tasks difficult
     
    #13     Feb 7, 2024
  4. mag2717

    mag2717

    Not really. A DL model is capable of doing much more than you wrote. What I mean is that a model can do more than just make predictions; it can actually manage your trading, handling all its aspects.
     
    #14     Feb 7, 2024
  5. MarkBrown

    MarkBrown

    like building a house the foundation is the most important - translated to trading terms that would be "data" if you don't master the data you have garbage.

    always the first failure point.
     
    #15     Feb 12, 2024
    beginner66 likes this.
  6. vanzandt

    vanzandt

    All the AI and computer learning in the world won't replace common sense.

    Search "VZ's lemonade stand" definitive analysis of how stocks work beyond the myopic quarter to quarter approach geared for idiots and pumped by popular websites and many of the folks on CNBC. You can either lose money, or invest in what's real for gradual long term wealth building. Water always seeks its level over time... that, and algos will always beat 90% of retail traders near term. Write that down.
    ~vz
     
    #16     Feb 13, 2024
  7. Quanto

    Quanto

    I agree with your first sentence regarding common sense, and the water analogy.
    But I give a shaet to such statistics comparing retail to others like professionals and algos. These stats can't be true, there is an agenda/intent behind spreading such IMO false stats, simply b/c it's very unnatural since it violates the Normal Distribution... :)
    Maybe someone wanted his group to stand out over others...
     
    Last edited: Feb 14, 2024
    #17     Feb 14, 2024
  8. About water always seeking its level, Soros would disagree (Reflexivity).
     
    #18     Feb 14, 2024
  9. vanzandt

    vanzandt

    Well, when I say "90% of retail traders", obviously thats just a blanket statement used for convenience, but you get the idea. Trading day to day is very tough, to win consistently takes time to learn. But it can be done. The #1 biggest thing... is always using small positions no matter how convinced you are you're right about a stocks move. That's easier said than done. That's why so many newbies zero out their accounts.
     
    #19     Feb 14, 2024
    Quanto likes this.
  10. Quanto

    Quanto

    Of course I agree. And it's not hard to do so in practice:
    by doing many smaller trades one can even get much more fun... :)
     
    #20     Feb 14, 2024
    vanzandt likes this.