I have looked through this thread, and found a mention of your research in the RL topic. If you can help me not to re-invent the wheel by answering a couple of questions:
What way of learning have you tried? e.g. q-learning, double Q-learning
Have you implemented multi-agent learning? Are there positive effects?
Did you use reward features, and if so, which ones?
And finally, why did you stop moving in this direction?
I have looked through this thread, and found a mention of your research in the RL topic. I am trying to dig in this direction, and if you can help me not to re-invent the wheel, answering a couple of questions:
What way of learning have you tried? e.g. q-learning, double Q-learning
Have you implemented multi-agent learning? Are there positive effects?
Did you use reward features, and if so, which ones?
And finally, why did you stop moving in this direction?
In advance, thank you!
I used a simple reinforce. And this is almost no different from optimization. It just maximizes the total revord. And why do it when you can, for example, just mark up the graph like in the article. The end result is a hitch on the new data, just like in normal training. But the topic is too big to draw unambiguous conclusions. The main thing is to know what to optimize: if I want profit, it is clear where it is maximal. I've tried different methods, but ended up not understanding what kind of rewards are better.
Allow the use of cookies to log in to the MQL5.com website.
Please enable the necessary setting in your browser, otherwise you will not be able to log in.
You are signed in. Two-factor authentication has been disabled. Set it up again and scan a new QR code, or continue with two-factor authentication disabled.
I don't know... maybe it's my experience or maybe it's my drinking...)
but it seems to me that you're a dick...)
maybe it's because you're a scornful bourgeois.
Maybe it's because you're a despicable philistine
offensive
shame
oops )
What do you think about my question? Is it realistic to attach a stop and a take out there?
https://www.mql5.com/ru/forum/86386/page2248#comment_19779215
What do you think about my question? Is it realistic to attach a stop and a take?
https://www.mql5.com/ru/forum/86386/page2248#comment_19779215
Yes, you can specify the conditions
can be set directly in the terminal. I have no time to change it.
I have looked through this thread, and found a mention of your research in the RL topic. If you can help me not to re-invent the wheel by answering a couple of questions:
What way of learning have you tried? e.g. q-learning, double Q-learning
Have you implemented multi-agent learning? Are there positive effects?
Did you use reward features, and if so, which ones?
And finally, why did you stop moving in this direction?
Thank you in advance!
Yes, the conditions can be prescribed
can be written directly in the terminal. I don't have time yet
Questions are bombastic ....
It's interesting to hear the answers )))
I have looked through this thread, and found a mention of your research in the RL topic. I am trying to dig in this direction, and if you can help me not to re-invent the wheel, answering a couple of questions:
What way of learning have you tried? e.g. q-learning, double Q-learning
Have you implemented multi-agent learning? Are there positive effects?
Did you use reward features, and if so, which ones?
And finally, why did you stop moving in this direction?
In advance, thank you!
Question, how does machine learning help in forex trading?
Question, how does trolling on the forum help in forex trading?