Machine learning in trading: theory, models, practice and algo-trading - page 34

 
mytarmailS:

What predictors did you feed there?

As I see it, the predictors and the target are like a whole, everything should be interconnected there, it's very stupid to want to catch exact bounces by feeding a 200 moving average as predictors. It's like two different universes that don't overlap.

I have an algorithm that determines the predictive power of a predictor for a particular target variable. In short, it's oscillators and different increments. If a particular predictor has predictive ability for a particular target variable, it does not follow that it will have predictive ability for another target variable. Moreover, a predictor may have predictive ability in one window and not in another.

The algorithm works well. The predictors it selects do not lead to overfitting of the models.

PS

According to my algorithm, slips of any type have no predictive ability no matter how ridiculous it sounds.

 
SanSanych Fomenko:

I have an algorithm that determines the predictive power of a predictor for a particular target variable. In short, it is oscillators and different increments. If a particular predictor has predictive power for a particular target variable, it does not follow that it will have predictive power for another target variable. Moreover, a predictor may have predictive ability in one window and not in another.

SAR

According to my algorithm, slips of any type have no predictive ability no matter how ridiculous it sounds.

The algorithm works well. The predictors it selects do not lead to overfitting of the models.

Can I see how the algorithm works?
 
mytarmailS:
Can I see how the algorithm works? I'd like to see how it trades

He doesn't trade, he selects.

If you distill your set, I will give you my selection result. Other than that I'll give an error on some models. But your target variable is not quite clear - extremely unbalanced classes.

 
SanSanych Fomenko:

He doesn't trade, he selects.

If you distill your set, I will give you my selection result. Other than that I'll give an error on some models. But your target variable is not quite clear - extremely unbalanced classes.

Well, can you show me a trading algorithm that trades on selected attributes?

 
mytarmailS:

Well, and the trading algorithm that trades on the selected features can you show?

No. In my trading robot I am trying to improve some of its elements using R. At the moment I have replaced trend forecast on high TF with rf prediction. The drawdown decreased and became flatter. But the profit factor has not increased. I was going to upgrade other components, but I ran into rottenness in TA. I'm facing a choice: to throw out all profitable trading robot and write everything in R, or to continue upgrading...
 
SanSanych Fomenko:
No. In the robot I have trading.....
Why not?
 
mytarmailS:
why not?

There is no particular need. The robot seems to be there, it brings some money....

R is more for the soul, yes fear that the robot will go stale as it was more than once - I do not manage to write on TA a robot that lives more than half a year. This one has already survived six months, maybe for a set of rf inserts, or maybe just lucky....

 
SanSanych Fomenko:

There is no particular need. The robot seems to be there, it brings some money....

R is more for the soul, yes fear that the robot will go stale as it was more than once - I do not manage to write on TA a robot that lives more than half a year. This one has already survived six months, maybe for a set of rf inserts, or maybe just lucky....

I have an account with a broker that somehow messes up the strategies. There is for example an EA that has been working more or less adequately for a year on one broker. As soon as I traded for a month on account of another, "black" broker - EA stopped bringing profit on both of them at once - stoplosses became more frequent. I stopped trading on the "black" account - profit on the first broker returned to normal. The broker apparently trades large lots on my strategy, very strange, I have several times tested and always with such a result.

If a strategy that is known to be profitable suddenly goes bad, I would advise to try another broker.

 
SanSanych Fomenko:

There is no particular need. The robot seems to be there, it brings some money....

I ask you to make a screenshot of your robot's trading and that's it... i think i have a robot, i think it may not? ))
 
Dr.Trader:

The broker apparently trades large lots on my strategy

I so suspect that you are on the forex, there are no brokers on the forex and they do not trade, these are shops that work on the documents of bookmakers

p.s. what do you think of my suggestion on the selection of traits?

Reason: