Machine learning in trading: theory, models, practice and algo-trading - page 1872
![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
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 could at least take it and look for the fault line by line. At the transitions of the hours sometimes there are strong gaps, clearly the hours are not one day, displaced
may be because of holidays or sessions opened at different times (holidays or something else)
corridor to do and at the exit of the opposite edge to rebuild or wait and then rebuild. Shift because of the conditions then returns to the place? If it comes back, then you should look at it. If it doesn't come back, then something is wrong.
Well fto, almost combed all the framework, now I have to saw the grails
I'm trying to automate as much as possible. For example, I have an idea to spar trading rules in another mql function, for each cluster
For example, what strategies would you use for each cluster?
according to the type of conditions... if the price of some opening is higher/lower than something (average or deviation), then we do this or that, depending on the predicted cluster
only the 2nd hour is involved in trading, i.e. the right half of the chart
it is possible to generate the whole mql code of the bot
Cluster0 is more static, both in terms of outliers and average.
The remaining clusters are skewed to one side or the other, the average is also not stable.
From this we should proceed. Build a statistical model on Cluster0.
do a corridor and rebuild when the opposite edge comes out, or wait and then rebuild. The change from the conditions then returns to the place? If it does, you have to watch it, if it does not, then something is wrong.
It seems to be connected with the summer/winter time changeover... but it's not exact...
Ahahahah ))
there's a new stable boy.)
latent mamkogeb ))))
You should at least say hello first , ahahahah... pathetic...
Ahahahah ))
there's a new stable boy.)
latent mamkogeb ))))
You should at least say hello first , ahahahah... pathetic...
and friends love you ;))
)))
Ahahahah ))
there's a new stable boy.)
latent mamkogeb ))))
You should at least say hello first , ahahahah... pathetic...
sounds like a magician by the way he communicates
In short, I localized the problem... it is in the data skips, which are pulled through the api. I switched to another dc - there are problems in another place. Somewhere there is a shift, in random places.
I need to fix the logic for this
Does anyone network, how do you balance the number of epochs, batting size and laning rate?