Machine learning in trading: theory, models, practice and algo-trading - page 1340
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
Honestly, I'm here in the hope that here's your next post will be "there are the first results....", all your research they at least somehow applied in practical terms, if not then maybe all your work at the moment is a road to nowhere?
Strange way of thinking - I have not yet done what I had planned, so it's too early to bet on the trade. At present I test predictors on real trade with New Year's herbarium, catching bugs and collecting information. Then I will set the model of CatBoost with a small lot, I need to check the correctness and speed of working with real data. If my profit will be at least 10k for 2018 on the test sample, perhaps with combinations of models, then I will put it on the real market with a normal lot. I'm not sure if I'll be able to sell my deposit. The development and debugging is not quick.
You can save the model in json and parse it in mql, i.e. you'll have a ready to use bot without any extra files.)
The model can be saved in json and parsed in mql, so you'll have a ready to use bot without unnecessary files
Of course, the overshoot is in 3 lines, yeah...
The problem is in the interpreter, not the model.
And, splitting into 7 files to speed up the process, not sequential command execution.i.e., is there a question mark inside the interpreter or what?
I don't know what's inside the interpreter - maybe you can write it and tell me.
I suggest a direct way to create a model, without a python wrapper.
this is a slow and thorny path, the bare model is only needed for use
This is not a good way to explore, only through python.
Just my variant, I think, will be faster than python, also due to parallel application of models - you want to check.
I need it for model generation, for fiddling with individual model I can use other tools including python, R or MT5.
Put the source code and the principle of work properly, with a good description
So, maybe someone will be interested and not with a black box to interact
Well, I think not only programmers read this thread, and people who want to try the MO, but do not know how to approach it, this script to help them.
The programmer, I think, will cope with the code - it seems that there is nothing clever there, except for working with tables, which is not my thing.
Yeah, we read it...
The programmers here couldn't figure out Reshetov's code, let alone dig into some vague, unformalized ideas.
not enough imagination
So you haven't actually done anything and suggest we do everything from scratch based on your fantasies.
I personally do not offer you anything, but provide an opportunity for those who are interested.
I didn't do anything for you personally by that action, no doubt about it.
Especially for those who are interested in the code, I asked the author of the class for permission to publish the code with functions from the class, without the class itself, if he does not mind, I will publish it. Maybe someone will remake the class functions and the script will be publicly available.Maxim Dmitrievsky:
The pinnacle of abstractionism is mathematical formulas and abstract concepts, not mush in the mind of some particular dreamer
Not exactly. The top is multidimensional images, not formulas and not abstract notions.
But for the brain to be able to operate with such thought-images it needs light energy, without it there will be degeneration and rolling on formulas.
Okay, cap.
I know dark energy, never heard of light energy.
The structure of the world around us by quantum levels. For those who do not know where they are).
Well fuck you to the left, here we go.