Machine learning in trading: theory, models, practice and algo-trading - page 877

 
Ivan Gurov:
So it didn't work out))))

Why didn't it work? You could probably start in the middle. I confess, I didn't read it carefully myself from the beginning).

 
Yuriy Asaulenko:

Why didn't it work? You could probably start in the middle. Admittedly, I didn't read carefully from the beginning either.)

I read from the end.

And Mishka disappeared somewhere.

I always had a real signal, a real trade on NS.

 
Renat Akhtyamov:

I've been reading from the end.

And now Mishka has disappeared somewhere.

I don't know, I've never seen any.

I don't watch the signals at all, I'm not interested. I haven't seen any at all.)

Who is Mishka?

 
Yuriy Asaulenko:

I don't watch signals at all, I'm not interested. I haven't seen any at all).

Who is Mishka?

Mihail Marchukajtes
 
Renat Akhtyamov:
Mihail Marchukajtes

He seemed to talk about the NS in the beginning, and then somehow this topic died down. And then it was drained in a couple of days.

It seems that there was nothing useful there. Yes, and he was very unclear - I did not understand a thing). I must say, and read from fifth to tenth.

 
Yuriy Asaulenko:

He seemed to talk about the NS in the beginning, and then somehow this topic died down. And then it was drained in a couple of days.

It seems that there was nothing useful there. Yes, and he was very unclear - I did not understand a thing). I must say, and read from fifth to tenth.

And who could show me the results of real account trading?

Mishka only showed it, thus proving the inadequacy of trading using a neural network.

 

Version 13 of Darch is now available.
https://github.com/maddin79/darch-installation from here only for now.

The update via CRAN does not work, because the previous version was sent to the archive due to errors. Maybe a new fixed version will be restored soon.

 
Renat Akhtyamov:

And who here can show the result of trading on the real?

Mishka has just shown, thus proving the inconsistency of trading using a neural network.

He actually proved absolutely nothing.

If you can't solve a problem, that doesn't mean that it has no solution at all. Even more - it does not mean anything at all).

I have a different methodology of applying NS in general - it is partially laid out in the thread - TC on neural networks (a very brief guide).

In brief, NS supplements the classical strategy, and serves as a teachable logic for decision-making in classical strategies. This allows to significantly simplify both the strategy and the NS itself.

 

Yuriy Asaulenko:

In a nutshell, NS supplements the classical strategy, and serves as a teachable decision logic in classical strategies. This allows you to significantly simplify both the strategy itself and the NS itself.

IMHO, this is the only use of NS in trading today. All others are a waste of time and effort. Given the current level of so-called AI))

 
Yuriy Asaulenko:

He actually proved absolutely nothing.

If you can't solve a problem, that doesn't mean it has no solution at all. Even more - it doesn't mean anything at all).

I have a different methodology of applying NS in general - it is partially laid out in the thread - TC on neural networks (a very brief guide).

In brief, NS supplements the classical strategy, and serves as a teachable logic for decision-making in classical strategies. This allows to significantly simplify both the strategy and the NS itself.

It turns out that the NS is a sort of filter for making decisions about entering a trade, a preliminary test of a possible result?

Operational tester, so to speak?