Machine learning in trading: theory, models, practice and algo-trading - page 3014
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
The person is not able to create an environment/cluster to run the same specific libraries, as I understand it.
So why is this reference in response to the non-universalism claim?
I give up,
You should be kinder and more constructive :)
If you claim that any code can be transferred for execution on another computer that is not clustered, then show it on a concrete example - my code.
Then I will be happy and grateful.
That's the way to be kinder and more constructive :)
If you claim that any code can be moved to another computer not clustered together for execution, then show it on a concrete example - my code.
Then I will be happy and grateful.
100$
Great, but with minor tweaks to the current code that I described above.
Should I open a freelance job?
Great, but with minor tweaks to the current code that I described above.
Should I open a freelance job?
No modifications.
Why?
About the leaves, here's dug up an old EA, culled on data from 2014 to 2019. May have discarded something in 2020 (can't remember), so tested on 2021 + month 2022.
Here is the situation on the profit of each sheet. The average value is -14 rubles. 48% shows a positive result.
Graph of the best leaf in 2021
Below is the balance of leaves in the order of their addition.
You can see obvious trends, which most likely indicates an error in choosing the root predictor when building the tree.
Let me remind you that there are 100000 generations of genetics, and each one builds/varies the tree. Then I remove the root predictor from the sample and repeat the process. If memory serves, I have run about 100 such cycles. It is not possible to quickly determine from which training cycle the trends emerged now, but this information can be obtained later - for reflection.
Now I saw that I had 4 more sheets in the code as filters to each signal sheet :) So I switched them off and now I am repeating the charts.
Here is the situation on the profit of each sheet. The average value is +34 rubles. 50% shows a positive result.
The best by profit
And the balance of all leaves
The situation here looks better, although there are also trends and the reason is probably previously mentioned.
It turns out that the leaves were not very qualitatively selected for filtering.
Like one sheet of wood sometimes works better than wood?
ready tc at minimum
Like one sheet of wood sometimes works better than wood?
ready tc at minimum
On the new data, yes. It's just that a consistent pattern has been hit, and the rest of the leaves from the tree are just wind noise. The noise is there from forcing the tree to give answers to all questions, i.e. to classify the sample completely.