Machine learning in trading: theory, models, practice and algo-trading - page 1241
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
I don't understand Accuracy either. But error matrix or classification error is more understandable, that's what I'm displaying on my screen.
Akurasu on unbalanced classes can show 100% class separation, but it will not be so. And the data will just all go to the majority class.
I have already installed it, tomorrow I will try it, along with just GBM, maybe LightGbm... a xgboost is something painful to configure, it takes a long time to figure out
More like a selection of parameters... which are more suitable for BP
Akurasu on unbalanced classes can show 100% class separation, but it will not be so. And the data will just all go to the majority class.
Wizard - what did Accuracy like?
It's more like a selection of parameters... which are more suitable for BP
maybe, i don't know) i also need to master mustache boxes in python and preprocessing, it's not too hard
Well, smart sample splitting is half of success so to speak
That's what I wanted to ask if anyone knows. By what criteria they select programmers in hedge funds and banks for quantitative analysis, mo and other things.
I do not understand who comes up with their algorithms the programmers or individual people, and then the programmers embody in the code?
That's what I wanted to ask if anyone knows. By what criteria they select programmers in hedge funds and banks for quantitative analysis, mo and other things.
I don't know who thinks up algorithms or individual programmers, and then the programmers implement them in the code?
It's a long time to tune everything properly, especially the data))) In short, don't chase parrots, and do the trend oos almost identically, then preprocessing. Doc learned how to do the trend and OOS almost 1 in 1, but the data to prepare and where to get it did not catch up...
Yeah, he wrote something that if it weren't for the spread... )
By family ties with the bosses )))) And 1 or 2 normal progectors to work for the entire department. But I don't know how they select the normal ones.
No, it doesn't work like that. Being a workhorse, we've been there....
It's not like that everywhere, is it?
No, it doesn't work like that. Being a workhorse, we've been there....
Not everywhere is like this, right?
Normal progres will be 100% leveraged in any case. Well, the main thing that the salary will suit)