Machine learning in trading: theory, models, practice and algo-trading - page 2515

 

Mda.... it's not very constructive at all. Only reproaches and trolling))

This thread is intended to discuss ideas for the use of MI, but no one is going to share their experiences (for obvious reasons).

Perhaps this is the apogee of this branch)))

 
iwelimorn #:

Mda.... it's not very constructive at all. Only reproaches and trolling))

This thread is intended to discuss ideas for the use of MI, but no one is going to share their experiences (for obvious reasons).

Perhaps this is the apogee of this thread)))

Alexander_K has arrived, now there will be constructive.
Alexander_K
Alexander_K
  • www.mql5.com
Профиль трейдера
 
iwelimorn #:

Mda.... it's not very constructive at all. Only reproaches and trolling))

This thread is intended to discuss ideas for the use of MI, but no one is going to share their experiences (for obvious reasons).

This may be the apogee of this thread)))

What's to share? There's Tinkoff shared...
is suitable for analysis and prediction of any processes - from the amount of precipitation in the coming winter to the company's needs for hiring new employees.

The modern MO cannot learn from market data. At least thank you for precipitation))

 
Aleksey Nikolayev #:

You already know how to calculate the ACF, don't you? Unlike smartlab, you can't ban me for this question here.)

The answer will be the same: I do not need to calculate the ACF SB. And even if someone does, I see no practical sense in it. I have only been granted one knowledge and that is enough. Who did? The Sorcerer, of course. One should be able to listen, and not try to calculate insanely.

 

Why don't MO amateurs get anywhere?

Yes, simply because you have to do preprocessing of the data and, of course, that data is not OHLC M1. Time! That's what needs to be worked on. It's been said a billion times, but only Max Dmitrievsky has explored something in this direction with unknown results.

 
Alexander_K #:

Why don't MO amateurs get anywhere?

Yes, simply because you have to do preprocessing of the data and, of course, that data is not OHLC M1. Time! That's what needs to be worked on. It's been talked about a billion times, but only Max Dmitrievsky researched something in this direction with unknown result.

Why, with a known, just like yours, zero
 
Vladimir Baskakov #:
Why, with a known, just like yours, zero

Hehe... How do you know my results, Vova? First of all you should say hello or something....

 
Aleksey Nikolayev #:

You already know how to calculate the ACF, don't you? Unlike smartlab, you can't ban me for this question here)

you haven't done the math yourself and no one will do it for you... if you can't even decipher your own letters... and you've been laughing at your acf for 50 pages (if not your entire pseudo-trading life)... - that's how you recognize/differentiate between the flunkies and the traders... the first ones are stuck in their nonsense, and for a long time, (trying to look smart with one word for life and a pack of henchmen who dream to stink all their life, whether smartly or stupidly)...

p.s.

and if you have created a model that is essentially your nonsense, then of course there will be autocorrelation - this is exactly the sign of a Poor Model..."Horses" also thought that they would laugh for a long time and not have enough intelligence to shut up (the latter statistically proved to be true)

 
Alexander_K #:

Hehe... How do you know my results, Vova? First of all you should say hello or something....

Hi! I put you on my list by the way, congratulations!
 
Vladimir Baskakov #:
Hi! By the way, I put you on the list, congratulations!

Hi. On the list? Oh, yes!!! I missed that very much. It's kind of irrelevant to the DOD, though. What are you doing here?