Machine learning in trading: theory, models, practice and algo-trading - page 3119

New comment
 
Forester #:

That's the kind of thing I'm talking about. I got it on the first screen I saw. Yesterday between 3:00 and 4:00 p.m.

new avatar's ready.

;)

 
Forester #:
No, I have not researched this topic specifically. I remembered a quick fall after a slow growth in the times when I tried to trade manually.
Perhaps it was remembered on emotions because it drained my deposit. I do not exclude that everything is even or even vice versa)))
In currencies, the difference can be between unequally valued currencies. But between more or less equivalent over a long period of time is not. Besides, it is possible to reverse the quote. Stocks and indices are different).
 
Evgeni Gavrilovi #:

Catboost has it.

model.get_feature_importance(type=catboost.EFstrType.Interaction)

I didn't see a way to specify the relationship of predictors (input data). Can you be more precise?


The importance of predictors is nothing at all to talk about in ANY models.

 
Forester #:

That's the kind of thing I'm talking about. I got it on the first screen I saw. Yesterday between 3:00 and 4:00 p.m.

Here's the reverse, an explosive rise after a sluggish fall.....

or a closer approximation of m1.

So it's all subjective.

 

who knows about syntactic parsers and BNF forms(Backus-Naura Form)

please let me know

 

I have a brilliant idea.

We take two slippery averages, one called the bear and one called the bull.

We make the bear blue, the bull brown.

So as not to confuse them, we chase this market up and down until we achieve fabulous results.

Polina.Zlotova.

Using tick. And only the tick one from MT5 deals analyser.

 
Lorarica #:

I have a brilliant idea.

We take two slippery averages, one called the bear and one called the bull.

Make the bear blue and the bull brown.

So as not to confuse them, we race this market up and down until we achieve fabulous results.

Polina.Zlotova.

Using tick. And only the tick one from MT5 deals analyser.

Don't take slippery ones, take sliding ones.

 
СанСаныч Фоменко #:

I didn't see a way to specify the relationship of predictors (input data). Can you be more specific?

Interaction is a relationship.


 
Evgeni Gavrilovi #:

Interaction is what a relationship is.


Thank you!

What does the "importance" of the predictors have to do with the "importance" of the predictors, which we get as the RESULT of the calculations? I understood your post about the interrelationship of predictors as a condition for calculation.

The importance of predictors is an extremely common result of model fitting. Many times I have tried to use "importance" to screen out "unimportant" predictors to reduce classification error. If you remove any predictor, the "importance" of the remaining ones is ALWAYS recalculated. I think the parameter you mentioned indicates importance-dependent predictors.

 
СанСаныч Фоменко #:

Thank you!

What does this have to do with the "importance" of the predictors" that we get as the RESULT of the calculations?

It is a part of the importance calculation.

In this case, we show bundles of two signs, the greatest weight is the relationship of two moving averages with period 74 and 137 for the prediction Close[i+1]-Close[i]

1...311231133114311531163117311831193120312131223123312431253126...3612
New comment