Shall I get some tips about useing Machine Learning on trading?

Discussion in 'Trading' started by cf0532, Dec 30, 2016.

  1. O(1)

    O(1)

    yea. I shouldn't have let it get under my skin. But, I really don't like when people start calling other people idiots. there's no need for that..

    my apologies. best to all and hey.. Happy New Year!
     
    #41     Dec 31, 2016
    VPhantom likes this.
  2. The "real" market in not random because you have to include "......" to make it not random. I will not go any further on this, I am not here to give free lunch.

    You still loss if you buy and sell random in long term, after you include the slippage and commission.
     
    #42     Dec 31, 2016
  3. The holy grail in trading is trading the most liquid thing, using leverage/options. for explosive atom bomb compounding gains.
    ...That's one half of it, the second half is the skill to pull it off o_Oo_O for the longer haul.
     
    #43     Dec 31, 2016
  4. or explosive of account blow up using leverage such as Future or options. :)
     
    #44     Dec 31, 2016
  5. algofy

    algofy

    Galvin you are such a genius, there is no other quite like you on these boards. I love how you say...."I'm right, you're wrong and I'm not going to waste any of my valuable time explaining it because I'm better than you. "

    This sums up 90% of your posts.
     
    #45     Dec 31, 2016
    VPhantom likes this.
  6. No one is better than anyone else. We all equal.
     
    #46     Dec 31, 2016
    algofy likes this.
  7. It doesn't take a quant or math whiz to figure this out, and you sure as fuck don't need machine learning to implement a robust algorithm that churns out profits. My models use the most basic math calculations applied to the right instruments, at the right time.
     
    #47     Dec 31, 2016
  8. ET180

    ET180

    You're throwing out a bunch of unsubstantiated generalizations. I never said that all programmers who go it alone will turn out to be rock stars. I just said that if one really wanted to learn it, they can. Good programming skills are mostly learned through experience anyway and most importantly the drive to learn and continuously become a more efficient programmer. Companies such as Amazon and Google recognized this as they have dropped a formal degree requirement for many of their jobs. Basically, if you can pass the programming interview, you're in.

    I can tell you as a software engineer at a major tech company that there's a big difference between practicing open heart surgery and software engineering. With software, you can make mistakes and learn from your mistakes. You can't afford to make mistakes when performing open heart surgery. A good software engineer recognizes their imperfection and develops tests to validate their software. This is something that can be learned independently and there's many great resources available on the Internet. I can tell you that if you watch, do the homework assignments, and really understand the youtube series CS106A, CS106B, and CS107 from Stanford, you'll be a competitive software engineer for an entry level software engineering job at any major tech company.
     
    #48     Dec 31, 2016
    aex and VPhantom like this.
  9. ET180

    ET180

    The market is not random simply because sometimes people buy or sell not because they want to or believe that the asset will go in their direction of preference. They buy or sell because they have to...they got caught in a bad trade and have to unwind their position. It's those opportunities of market inefficiency (take a look at XBI a few weeks back, it was 60.xx one day, 63 the next, and then back again a day or two later...that wasn't the election, happened sometime in December). That's not a characteristic of an efficient market. There's no reason a large biotech ETF with lots of diverse holdings should experience a 4 or 5% repricing from one day to the next on no news. It's fear and greed that drives the market. These periods of inefficiency and definite non-randomness provide opportunities to profit.
     
    #49     Dec 31, 2016
    VPhantom likes this.
  10. ironchef

    ironchef

    I enjoyed reading your posts.

    Well, the truth is sometimes very painful. Been there done that. I gave up day trading a long time ago and let true professionals played in that sandlot.

    I did this calculation once: Plotted 1 day price change of SPY, overlaid with a set of randomly generated number and the two were almost identical. If I looked very very closely perhaps I could see some differences but I was not sure it was just sample size errors. I am just not smart enough to figure out the small non random part of SPY, if it exists.

    I do have a question for you. I think the long term stock prices are not random (they do have trend lines), so if short term, intraday and days they are random, when does the transition occur from random to a trend? And what happen at those transitions?

    Best wishes.
     
    #50     Dec 31, 2016