Machine learning in trading: theory, models, practice and algo-trading - page 572
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Divide a large deposit into smaller ones. You trade on one account and copy trades on the others.)
Genius.
how to fit models :))
current increments and bottom trained model, 2 trees :) if I set 20 trees, then 1 to 1
And the input current increments of the pair are not fed to the input, and the model is trained on them (fed to the output)
For some reason it remembers it perfectly, now I need to predict the future
how to fit models :))
current increments and the bottom trained model, 2 trees :) if I put 20 trees, it is 1 in 1
For some reason it remembers perfectly, now I need to predict the future.
It remembers really well. The only question is, do the increments have anything to do with the future? I can not claim, but it seems to me that they do not.
By the way, I don't understand why the flat nodes have such big increments, or do you mean something else by increments?
Memories are really great. The only question is, do increments have anything to do with the future? I can't say for sure, but it seems to me that they don't.
By the way, I don't understand why the flat vertices have such big increments, or do you mean something else by increments?
here increments with a lag, for example 55... that's why they are not distributed so noramally, that's why flat vertices can be large
well, let's see what can be squeezed out of it... i don't know if it has a future or not, reserch should be done
i'm just surprised that the forests are so fit... i can't find the error for half a day now, as if everything is really right :)
Here the increments have a lag, for example 50... that's why they are distributed so abnormally
Well, let's see what can be squeezed out of it... I don't know if they have it or not, resurvey should be done
Is this a sort of momentum?
Well, not the increments of the adjacent bars are taken, but for example close[0]/close[55].
how to fit models :))
current increments and bottom trained model, 2 trees :) if I set 20 trees, then 1 to 1
and the input current increments of the pair are not fed to the input, and the model is trained on them (fed to the output)
It memorizes it perfectly for some reason, now I need to predict the future
I spent half a year with different predictors, including the increments of anything to anything, the number of predictors. And I used different models. And RF, and SVM, and MLP. I even tried the higher times and went all the way down to M1. The maximum I could achieve with a validation sample is 53% of accuracy, and with the training sample - 100.0%. This is not enough for trading. In order to be in the black, I need at least 57% of accuracy. I don't know if it's my hands or what. Has anyone achieved better results? I am just curious.
Tortured for half a year with different predictors, including the increments of anything to anything, the number of predictors. And I used different models. And RF, and SVM, and MLP... I even tried the higher times and went all the way down to M1. The maximum I could achieve with a validation sample is 53% accuracy; the maximum accuracy for training is 100.0%. This is not enough for trading. In order to be in the black, I need at least 57% of accuracy. I don't know if it's my hands or what. Has anyone achieved better results? I'm just curious.
I haven't reached it that way yet... I'm making different variants... I have to look through a lot of variants.
Well, the increments are taken not by adjacent bars, but for example close[0]/close[55].