Can deep learning predict the stock market?




[ad_1]

Check out my follow-up video where I explain how some financial market crashes can be predicted:
https://youtu.be/6×4-GcIFDlM

Source


[ad_2]

Comment List

  • Leo Isikdogan
    January 19, 2021

    So should I suspect that AI technology is currently unable to trade stock or currency markets and make profits while being reliable, I'm a good programmer in mql4 and I admit that no strategy is going to work efficiently and reliably every day and on all pairs but I had so much hope in Machine Learning, Actually I was planning to study it just to be able to program a trading AI, you make me lose hope now.

  • Leo Isikdogan
    January 19, 2021

    Can you expand more on what you mean by picking random subset of stocks at every epoch ? Does this mean all the mini-batches in an epoch will have same subset of stocks but the subset of stocks will be different from epoch to epoch ?

  • Leo Isikdogan
    January 19, 2021

    another big talker no action or showing these utubers are the worst!

  • Leo Isikdogan
    January 19, 2021

    It took me a while to realise his voice wasn’t a joke

  • Leo Isikdogan
    January 19, 2021

    It's well known that predicting price movements is impossible: while you may be right most of the time, you can be extremely far the few times you're not. It doesn't mean trading can't be profitable, only that methods based on being right about price movements are simply not how one should go about it.

  • Leo Isikdogan
    January 19, 2021

    What's wrong with your voice?

  • Leo Isikdogan
    January 19, 2021

    So, Lee sedol is now interested in A.I.

  • Leo Isikdogan
    January 19, 2021

    Did you make a model that can make profit? I'm going to do the same thing like you said, is this a bright way to go?

  • Leo Isikdogan
    January 19, 2021

    shut uuuuup ! annoying voice dude

  • Leo Isikdogan
    January 19, 2021

    you do sound like Mikey Mouse. Sorry, but I have been laughing so hard while listening this. hahahahahhahahaha

  • Leo Isikdogan
    January 19, 2021

    EMH is a convenient starting point.. that's all.. nothing more, nothing less

  • Leo Isikdogan
    January 19, 2021

    I may not be an expert, but what about the human factors? Fear? Greed? They are the movers in the stock market … Yes, there are algorithms and they are probably the fuel added to the fire but we usually forget that the charts are a game of emotions playing up, fx: people tend to leave their stop losses to be reached as they are hopeful it will change whereas some traders are quite impatient and take their profits way before they should have – "better this than nothing".

  • Leo Isikdogan
    January 19, 2021

    I like how ur thumbnail looks like the doge meme

  • Leo Isikdogan
    January 19, 2021

    Ur voice was🤣🤣

  • Leo Isikdogan
    January 19, 2021

    You’ve inhaled too much helium.

  • Leo Isikdogan
    January 19, 2021

    in other words, it's useless and this video is a waste of time

  • Leo Isikdogan
    January 19, 2021

    📌 Second thought: some financial bubbles can be predicted. I explained how in my follow-up video:
    https://youtu.be/6×4-GcIFDlM
    P.S. My voice doesn't crack in the new video 😄

  • Leo Isikdogan
    January 19, 2021

    Voice of narrator too high pitch.. hard to listen

  • Leo Isikdogan
    January 19, 2021

    This dude sounds like Lee Sedol

  • Leo Isikdogan
    January 19, 2021

    I loved it . Thanks a lot . greatly appreciate it

  • Leo Isikdogan
    January 19, 2021

    Good informations, but there a lot of proofs the market couldn't be efficent at all. Like first example, a large company want to buy large amount of stocks, but didn't the price rising to high… If the market is efficent, there is no strategy to buy it for lower price for longer time. Most important, most forget – It is not about for what price you buy a stock, but for which you want to sell it and how much you buy and this are the more difficult questions. Stocks not go up forever you have to defne a point you want to sell and what to do, if you bought the wrong stock, when do you know after 10% drop, 15% etc.

  • Leo Isikdogan
    January 19, 2021

    There was a company that counted cars in a McDonalds car parks using cubesats

  • Leo Isikdogan
    January 19, 2021

    you need to have your servers close to the exchange because picoseconds matter

  • Leo Isikdogan
    January 19, 2021

    Your voice man

  • Leo Isikdogan
    January 19, 2021

    I learned that 10 yrs ago… exactly that… i learned machine learning like yousaid and matlab… anyway thankyou for remember me…

  • Leo Isikdogan
    January 19, 2021

    Got it. ML cannot help me to get rich.
    So, instead I can build a bot to give me the news about any stock I am interested in and buy it before its boom or sale it before the dip, right?

  • Leo Isikdogan
    January 19, 2021

    Uhm, your view is entirely cynical and doesn't even attempt to actually implement any machine learning predictions.. which can actually be fairly accurate and a useful tool for making decisions. It also seems your understanding of AI is limited and misguided..

Write a comment