Machine learning in trading: theory, models, practice and algo-trading - page 547
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 think if you connect Python properly, through the DLL will be faster. I've decided to write a header file in MQL5 to connect Python. I have decided to write a header file in MQL5 to connect Python. I will post the code on GitHub.
I will use RAMdisk or csd, no problem... The speed loss when restarting the script is a loss of seconds, which in the tester will be very slow... If you call it once, then it will be OK... + still need to save the model you trained so you don't have to retrain every time
By the way, https://www.mql5.com/ru/forum/223473/page9#comment_6279990
That's the thing: all loading and initialization is done through DLL in the MQL Init and DeInit functions, and the script itself calls functions from the Python code. This will be faster. Although in the tester it will be slower.
I'd rather put it on GitHub.
In ml-assistant I call the external script. It's true, it takes a few seconds to start the shell and loading the libraries, but it's not a big deal for me, since I'm not going to go lower than m15-m30. Taking into account the fact that the prediction on a trained model takes milliseconds, it's not that bad with a good model.
As for the tester, I didn't even try to run it with a trading robot, I attached "trading" to the chart at once for debugging.
I even found a vds server with acceptable power and costs and running on ssd.
I have found a virtual trading server on the server that is acceptable in terms of power and costs. In section M A R K E T O P E R A T I O N S
By the way, there are whole testers and resources for python where you can get online reports, if interested I will send you the links... I have not used them myself, just googled something to read
I found this about the reports: https://pypi.python.org/pypi/trackml
But I haven't figured it out yet. For some reason, the official site does not allow me to go.
I found thishttps://www.quantopian.com/
and here is the site of a guy who teaches python, including for financehttps://pythonprogramming.net/getting-stock-prices-python-programming-for-finance/
I like it, I think I'll buy a T-shirt and a cap with the python logo on it later for support :)
I found thishttps://www.quantopian.com/
and here is the site of a guy who teaches python, including for financehttps://pythonprogramming.net/getting-stock-prices-python-programming-for-finance/
I like it, I think I will buy a T-shirt and a cap with the python logo to support him :)
Here Google offers free computing power of its servers.
I have heard about quantopian, but did not understand what their main idea is. Now I checked it, it looks like Google's service https://towardsdatascience.com/neural-networks-with-google-colaboratory-artificial-intelligence-getting-started-713b5eb07f14.
Well, there's a tester for strategists
google lab seems to be awesome
I will support the topic. I use Amazon services, but their model builder is not so good. In any case, I could not build more or less5 quality model. Although maybe I did something wrong, but there are not too many settings there. Now I will try it on Google...