Machine learning in trading: theory, models, practice and algo-trading - page 2095
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 do not know anything, but I will support you.
There's a reason for that. It's an evaluation of the FA. Which can be used for any purpose and tasks.
https://habr.com/ru/company/skillfactory/blog/526542/
and maybe someone else will be interested (it looks like I'm going to get sucked into Python)
A quote - "Google Colaboratory is a relatively recent cloud-based service from Google aimed primarily at simplifying machine learning research and providing everything you need to do it right in your browser.
Using Google Colaboratory, you can remotely access a machine equipped with a powerful nVidia Tesla K80 GPU, and for free. This makes life a lot easier when it comes to doing experiments related to machine learning.
With Google Colaboratory you can:
- master programming in Python
- learn to work with popular packages for building deep learning models, such as TensorFlow, Keras, PyTorch. If for some reason the package you need is missing, you can easily install it using the package manager pip
Google Colaboratory is a Jupyter Notebook platform (more about Notebook here: https://webdevblog.ru/jupyter-notebook-dlya-nachinajushhih-uchebnik/) , deployed in the cloud. It allows you to run Jupyter Notebooks on a remote machine. These notebooks are .ipynb files stored in Google Drive."
It's good stuff, go for it.
If the data is not too big, then it is more convenient on the local PC. In google colab kernels are deactivated at idle, you have to reinstall all packages and download data. The processor there is only 4 cores and tesla is not the most powerful
they probably upgraded the cores to one ) if you want more you have to pay for it
I don't get it, is text more convenient than ssv?
What's the difference? I just read from it and that's it
Have you tried putting all the currencies?
Maxim Dmitrievsky:
If there is a critical mass of python users here - they will add native support for ONNX to Metatrader, it will be easy to transfer models from python directly to the terminal. Test it, publish it in the marketplace and so on. But so far everything is very dull and lifeless... or put + if there is someone alive who needs it or is interested. Let's say there is a good trained network on tensor or torch showing results. It is almost unrealistic to transfer it to metatrader. The same catbust supports onnx.
(I agree, although I still have very little idea what ONNX is )))) but I have had a quick look.
There is a kind of working model in keras and it seems to work on my testers. Now I am trying to run it in Metatrader on a demo account. Maybe it will be possible to connect it to ONNX
What's the difference? I just read from it, that's all.
Didn't you try stuffing all the currencies?
Do I look like an idiot? )
I agree, although I do not understand what ONNX is yet ))) but I got acquainted with it in a cursory glance.
There is a kind of working model in keras and it seems to work on my testers. Now I am trying to run it in Metatrader on a demo account. Maybe it will be possible to connect it to ONNX
This is something like a json for saving and transferring models, a common format. Can be transferred to other languages and from environment to environment
https://www.mql5.com/ru/forum/306688/page78#comment_19128403
https://www.mql5.com/ru/forum/306688/page77#comment_19128330
this is something like json for saving and transferring models, a common format. You can transfer to other languages and from environment to environment
https://www.mql5.com/ru/forum/306688/page78#comment_19128403
https://www.mql5.com/ru/forum/306688/page77#comment_19128330
Do I look like an idiot? )
why?
Why?
I'm at a roll-up conference, I'll post later