Machine learning in trading: theory, models, practice and algo-trading - page 900
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
Is the grail ready yet? With the AI)
well that's it... you're already oligarchs ))
Of course, Maximko always achieves his goal.
of course, Maksimko always achieves his goal
)))
That's it... We abandon all our robots, get in line for the wonder-intelligent robot.
)))
That's it... We're leaving all our robots behind and getting in line for the wonder-intelligent robot.
I'm already halfway to the Swiss bunker with paramilitary guards, following Alexander with his dusty sacks full of bills, don't mention it!
Alexander_K2 from another thread
I would like to connect mt5 with python to be able to visualize it and check it immediately... now I am considering the possibility to call python shell directly through winapi and send commands from bot to python, I don't know if it is possible. R and dll do not digest and do not know how and do not want to work with them, although python is not required (looking at articles and works of old-timers) less and less desire to get into the thick of it - 500000 packets and the output is the same as from the bot on 2 MA
Link to such a monster would be useful to many, but not many under its power.
If I wanted to use 10 MAs with 16 steps and make a predictor - for each one - the price opened above/below MA and one more - number of MAs when numbering all MAs from the top to the bottom, it would be a model that describes the market?
Did a little experiment with dropping whole models (10 models in total are given). Forward from 2018.04.01.
Without dropout:
The following numbers mark how many random models were left, the rest were dropped:
Well some kind of this is what it turns out to be. Probably not very revealing, since the model turned out pretty good without dropouts. And the models are too similar to each other, which is bad, you need to revise the learning algorithm (game theory to help). But still some flexibility in the choices appeared.
It looks good! Stop at 1 (1 model).
I have to be able to see how many people are interested in this kind of monster, but not many.
As for the MAs, just falling asleep I was thinking, what if we take 10 MAs with 16 steps and make a predictor - for each - the price opened above/below the MA and another - the number of MAs when numbering all MAs from the top of the chart to the bottom, will such a model describe the market?
That's no way to put the question. Experiments, experiments... ...because nothing is obvious in theory. I've tried about a hundred different variants of TS, and it's scary to think about it.
The branch history keeps a lot of information on how not to do it (and sometimes how to do it) is quite useful.
I have to be able to see how many people are interested in this kind of monster, but not many.
What about MAs, I was just thinking while falling asleep - what if I take 10 arrows with 16 steps and make a predictor - for each one - the price opened above/below MA and one more - number of MA when numbering all MAs from the top to the bottom, will such a model describe the market?
There was an article about 10 years ago with such an experiment with MA from 2 to 100.
And I think it's called a fan, if I'm not mistaken...
It seems to solve all the problems of mql4 communication with different languages. There is even a code for R. Here is the schematic.
The whole description is in three parts:
https://blog.darwinex.com/zeromq-interface-python-r-metatrader4/
https://blog.darwinex.com/zeromq-trade-execution-metatrader-zmq2/
https://blog.darwinex.com/zeromq-transaction-reporting-metatrader-zmq3/
Approved:
Why ZeroMQ?
Enables programmers to connect any code to any other code, in a number of ways.
2.Eliminates aMetaTrader user's dependency on just MetaTrader-supported technology (features, indicators, language constructs, libraries, etc.).
Traders can develop indicators and strategies in C/C#/C++, Python, R and Java (to name a few), and deploy to market via MetaTrader 4.
Leveragemachine learning toolskits in Python and R for complex data analysis and strategy development, while interfacing with MetaTrader 4 for trade execution and management.
ZeroMQ can be used as a high-performance transport layer in sophisticated, distributed trading systems otherwise difficult to implement in MQL.
Different strategy components can be built in different languages if required, and seamlessly talk to each other over TCP, in-process, inter-process or multicast protocols.
Multiple communication patterns and disconnected operation.
Here is the code
And here is the code for r