Machine learning in trading: theory, models, practice and algo-trading - page 861

New comment
 
Mihail Marchukajtes:

No, it's not. I call it a fucking microloan. That's it. Here's the seven of me. Starting a new account...

Oh, you're such a fucking idiot.

 
Mihail Marchukajtes:

No, it's not. I call it a fucking microloan. That's it. Here's the seven of me. Starting a new account...

And once again unavailable...

Make it public, if you don't mind.

 

Made (remade) a cool volatility indicator. I added volumes.

Does anyone need it as a feature?

https://www.mql5.com/ru/code/12567

for(bar=limit; bar>=0 && !IsStopped(); bar--)
     {
      residual=price[bar]-price[bar+1];
      residualSq=MathPow(residual,2.0);
      GarchNumVar=e_alpha*residualSq+e_beta*GarchNumVar;
      GarchNum=MathSqrt(GarchNumVar);
      long vol[];
      CopyTickVolume(_Symbol,0,bar,1,vol);
      IndBuffer[bar]=MathSqrt(GarchNum*vol[0]);
      //---
      if(bar) GarchNumVar_=GarchNumVar;
     }

I added tick volume and took the root from all this (I don't know why, but it defines cycles normally)

I marked in red where lag can be changed.

Homework: predict the volatility for n bars ahead and convert back to the price

The easiest ways: autoregression and https://www.mql5.com/ru/articles/318

 
Renat Akhtyamov:

And once again unavailable...

Make it public, if you don't mind.

What do you mean public???? Signal???

It's just that with 1:1000 leverage it's not possible to connect to the signal. I'll think about making it 1:100 like in the good old days.....

 
Vizard_:

Don't. Carry the title of honored teacher proudly!
Icarus must smack his face loudly on the pavement.
Then don't forget to fill the branch with snot. Have fun.)

Finally, I thought I was blind. I couldn't see your sarcastic jibes anymore..... Now it's okay, I'm relieved :-)

Well, don't worry about it... We'll get through this.

 
Maxim Dmitrievsky:

Made (remade) a cool volatility indicator. I added volumes.

Does anyone need it as a feature?

https://www.mql5.com/ru/code/12567

I added tick volume and took the root from all this (I don't know why, but it defines cycles normally)

I marked in red where lag can be changed.

Homework: predict the volatility for n bars ahead and convert back to the price

The easiest ways: autoregression and https://www.mql5.com/ru/articles/318

Volatility looks interesting. I think you can try to extract something out of this indicator. But the future volatility can be predicted by using the formula of Blek-Scholes. If you learn to predict the volatility, you can make a quick profit on options. Except that in MT options are kind of absent.... which is an omission....

Is there an option to make this indicator for MT4??? I think I'll be sitting with my models all night tonight, so I could include this indicator in my TS and see how it works out. Is there any sense in it.....

 
Mihail Marchukajtes:

Volatility looks interesting. I think it is possible to try to squeeze something out of this indicator. But the future volatility is predicted with the help of the Black-Scholes formula. If you learn to predict the volatility, you can make a quick profit on options. Except that in MT options are kind of absent.... which is an omission....

Is there an option to make this indicator for MT4??? I think I'll be sitting with my models all night tonight, so I could include this indicator in my TS and see how it works out. Does it make sense.....

I don't have time to rewrite it, I demolished mt4 more than half a year ago, it's not even installed on my computer. It's a simple formula, I think you can handle it. I don't know how to put it in a trading platform.

 
Maxim Dmitrievsky:

It's hard to rewrite it, I took MT4 down more than half a year ago, it's not even installed on my computer. It's a simple formula, I think you can do it. I don't know how to use it, I will try to put it in a different way.

Can you advise me to recommend any training course on Big Data? It should cover all aspects... May be in English)
 
Evgeny Gavrilov:
Any training course on Big Data can you advise?) It should cover all aspects... It can also be in English)

I'm looking now :)

https://yandexdataschool.ru/edu-process/courses/machine-learning

 
That's it... The account is activated. Now the statistics are perfect. Model received... Burn.... Hands in pants... Because of them (crooked hands), the statistics of the previous account was spoiled.... in general, waiting for the first signal and alga....
1...854855856857858859860861862863864865866867868...3699
New comment