Machine learning in trading: theory, models, practice and algo-trading - page 2776

 
Uladzimir Izerski #:

Especially for this thread, now for half an hour, using only OHLC, roughly sketched an arrow indicator.

Where is the indicator itself? If it is specially for this branch...

 
Aleksei Stepanenko #:

Where's the indicator itself? If it's specifically for the branch...

I can only show the result of work. An indicator will be too fat).

By the way, you had an opportunity, but you refused. Unfortunately.

f253jk

 
Maybe with machine learning the result would be more accurate. But I'm too lazy to bother with it. I'm waiting for an intelligent suggestion.
 
Uladzimir Izerski #:
Maybe with machine learning the result would be more accurate. But I'm too lazy to bother with it. I'm waiting for an intelligent suggestion.

Can you make a sample on the basis of your indicator on quotes from MQ on MT5 for 50 thousand bars?

Just record the date for synchronisation of each bar and, say, "1" - up arrow and "-1" - down arrow, "0" - no arrow?

And I'll try to train myself to guess the appearance of the arrow, and maybe even the direction. Let's see what the MO or my predictors are good for.....

 
Uladzimir Izerski #:

Even tried going in.

I hope it's on a demo, because trading under this is a bit of an exercise.

 

It's agood article, high quality.

higher than an overview but without revealing secrets

Machine learning for Trading: Adventures in Feature Selection
Machine learning for Trading: Adventures in Feature Selection
  • 2016.03.04
  • Kris Longmore
  • robotwealth.com
Back when I originally wrote this article, there was a commonly held idea that a newly-hyped approach to predictive modeling known as machine learning could discern predictive patterns in market data. A quick search on SSRN will turn up dozens of examples of heroic attempts at this very thing, many of which have been downloaded thousands of...
 
Aleksey Vyazmikin #:

Can you make a sample on the basis of your indicator on quotes from MQ on MT5 for 50 thousand bars?

Just record the date for synchronisation of each bar and, let's say, "1" - up arrow and "-1" - down arrow, "0" - no arrow?

And I'll try to train myself to guess the appearance of the arrow, and maybe even the direction. Let's see what the MO or my predictors are good for.....

So far it's a very crude version.

We'll see what happens in the future.

mytarmailS #:

hopefully on a demo, otherwise trading khat under this is a bit of a chore.

If you are sure, it doesn't matter what you trade on.

 

СанСаныч Фоменко #:

And so, all sorts of variations of something with something. I don't have a problem with that, making stuff up.

Could you publish one such predictor?

Maybe it's the difference in the RSI indicator reading on EURUSD and GBPUSD?

 
Uladzimir Izerski #:

An indicator would be too greasy.)

Vladimir, if something has worked out for you, I'm glad. So how does it relate to the branch? Neural networks or obtaining data for them?

 
Aleksei Stepanenko #:

Vladimir, if you got something, I'm glad. So how does it relate to the branch? Neural networks or getting data for them?

There's been a search for predictors here for a long time, 2775 pages. Something else to feed into the MoD.

My opinion is that it is necessary to search in the price itself, not in MAs of indicators and other distortions of the price.

I want to push those who suffer on the right path. They are stubborn).