Machine learning in trading: theory, models, practice and algo-trading - page 2687
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
You made a third party solution, protected it as an exe file.
Is it strange that you didn't give open source code in Python? You could have easily thrown a *.py script on the chart and everything would have worked?
That's why we put so much effort to make it possible to write complete and protected ML programmes within MQL5.
there is an open .py file attached to the bottom of the article, the .exe is made for those who don't want to install python and bother with library installations.
Support for running *.py files in the terminal is the ability to run already developed models inside the terminal. And the standard MetaTrader 5 for Python library allows normal communication and trading.
It is not for nothing that I pointed out earlier that all third-party integrations are a way of binding with duct tape with a predictable result. We are not interested in such paths at all. So asking to add duct tape is unnecessary, as well as skipping not a solid solution.
As you can see, the MQL5 language is growing very much and we, as its developers, know and publicly declare the goals. Then we achieve them.
Don't worry - almost everything in ML can be done in MQL5. This is exactly the goal we are heading towards.
Over the past 22 years, we have learnt how to achieve big goals.
I will not argue, MetaTrader is a high-level proprietary product, the boss is the boss. You asked a question, I expressed my opinion.
Maybe you are growing, maybe you are shrinking, it doesn't concern me, there is no fanaticism on my part to read and wait for updates. I just selfishly want to solve my problems and am looking for tools. I think that's the logic most people follow.
A page long instruction in small print "how to run a neural network in MT5" is not normal, and the impossibility to monetise it doesn't add enthusiasm either.
I have read Machine Learning, and some childish questions have arisen:
1. How to train a model? Where are the functions like in TensorFlow for forming layers and creating a model. Where is the analogue of model.fit(...) for training?
2. How to interrogate the model? Where is the analogue of model.predict(...) ?
These are the first one and a half questions, there are 125 more in the queue.
Maybe MT5 will come over to my side and I'll just keep doing what I've been doing.
MT5 will only ever be on your side, and you will keep doing crutches.
I read Machine Learning and had some childish questions:
1. How to train a model? Where are the functions like in TensorFlow for forming layers and creating a model. Where is the analogue of model.fit(...) for training?
2. How to interrogate the model? Where is the analogue of model.predict(...) ?
These are the first one and a half questions, there are 125 more in the queue.
They want to add compatibility via onnx at least partially, ML engine is not available yet
.
I don't get it, so you can do it in python and then turn it into .ex5 ?
I don't get it, so you can do it in python and then turn it into .ex5 ?
I looked into python integration and did not understand the logic of the creators.
You can get all the basic market information and can return a command to open a position to the terminal, but you cannot get a message from your own expert or indicator. There are so many functions for information exchange, and there is no such a small thing ))
So a pythonist sits down to write something using this bible and on the second step asks "what should I do with these bars and a lot of information about the terminal, symbol, last error, where are the indicators or my own expert????".
I wonder what kind of logic of use the creators of this biblical system had in mind.If we ask to add such a function, the answer will be about "duct tape"? Or we must be a dumb herd that doesn't understand anything.
There is clearly a dissonance between the pressure of progress and the market and the desire to keep their sandbox untouched. It seems that the development of the ML theme will follow the same logic.
Revisited the python integration and didn't understand the logic of the creators.
You can get all the basic market information and can return a command to open a position to the terminal, but you cannot get a message from your own expert or indicator. There are so many functions for information exchange, and there is no such a small thing ))
So a pythonist sits down to write something using this bible and on the second step asks "what should I do with these bars and a lot of information about the terminal, symbol, last error, where are the indicators or my own expert????".
I wonder what kind of logic of use the creators of this biblical system had in mind.If we ask to add such a function, the answer will be about "duct tape"? Or we are probably a dumb herd that does not understand anything.
There is clearly a dissonance between the pressure of progress and the market and the desire to keep their sandbox untouched. It seems that the development of the ML theme will follow the same logic.
The logic of introducing integration with python was to satisfy some specific requests of some very specific foundation. Closer integration clearly does not fit into the metaquote business model and so it won't happen, and trying to "teach them the business" will be futile and amusingly naive as always.
Following my advice will save you time and nerves.
The logic of introducing python integration was to satisfy some specific requests of some very specific foundation. Closer integration clearly does not fit into the metaquote business model and so it will not happen, and attempts to "teach them the business" will, as always, be futile and amusingly naive.
Following my advice will save you time and nerves.
The highlighted sentence sounds funny.
And do not attribute to me the desire to teach someone or ask for something. I'm just contemplating and giving my assessment. The fact that the project has already passed the peak of its development is obvious. The topic of crypto is failed, with the ML topic in the role of catching up, there is no normal browser version and, it seems, will not be.