What to feed to the input of the neural network? Your ideas... - page 16
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 know, I don't care what you think. You don't have your own neural network, but I do and I can tell my colleagues where the money is. Do NOT predict the price, but predict the probability of profitable trades.
You know, I don't care what you think. You don't have your own neural network, but I do and I can tell my colleagues where the money is. Do NOT predict the price, but predict the probability of profitable trades.
Being in this thread, please tell me what you feed to the neural network input. Especially since you have everything working.
Unless it is a trade secret.
Being in this thread, could you please tell me what you feed to the input of the neural network. Especially since you have everything working.
If not a trade secret
results of virtual trades on each bar, but it is not a secret - many people know about it...
results of virtual trades on every bar, but it's not a secret - many people know about it...
Thank you.
I mean, trading emulation. That's interesting
Thank you.
I mean, emulating a trade. Interesting
the main thing is on every bar.
most importantly, on every bar.
Tell me, do I understand you correctly:
1. I take any strategy, or compose the rules of the strategy myself
2. I trade it virtually, setting any virtual TP and SL (but judging by your screen, there is either TP = SL, or SL is one and a half times more than TP).
3. If on the current bar we reached the virtual TP - add +1 to the variable of profitable, if we reached SL - add +1 to the variable of losses.
4. Two values are fed to the neural network input.
If you have a much more complicated scheme, then this scheme - does it reflect the essence of successful implementation?
Tell me, do I understand you correctly:
1. I take any strategy, or make up the rules of the strategy myself
2. I trade it virtually, setting any virtual TP and SL (but judging by your screen, there is either TP = SL, or SL is one and a half times more than TP).
3. If on the current bar we reached the virtual TP - add +1 to the variable of profitable, if we reached SL - add +1 to the variable of losses.
4. Two values are fed to the neural network input.
If you have a much more complicated scheme, then this scheme - does it reflect the essence of successful implementation?
No not right, no strategies! We trade on each bar in both directions (BUY/SELL) and analyse...
No it's not right, no strategies! We trade on each bar in both directions (BUY/SELL) and analyse...
Unfortunately, quite a broad interpretation.
- If you open on the opening bar in both directions, the balance will stand at zero. Loc.
- If we occasionallyBUY, occasionallySELL, then we have a sign of a trading strategy .
- If there is no trading strategy, there is still opening in both directions, as in the first point, but with setting TP and SL.
Only in all variants there is no analysis, because there is no analysed object - trading strategy. Instead of analysing, there is a collection of statistics
Unfortunately, quite a broad interpretation.
- If you open on the bar opening in both directions, the balance will stand at zero. Loc.
- If to break off episodicBUY, episodicSELL, then we have a sign of trading strategy.
- If there is no trading strategy, we are left with opening in both directions, as in the first point, but with TP and SL setting.
Only in all variants there is no analysis, because there is no analysed subject - trading strategy. Instead of analysing, there is a collection of statistics
and you look at these trades separately, not in combination. Separately BUY and separately SELL. And the neural network will operate with these statistics.
and you look at these trades separately, not in combination. Separately BUY and separately SELL. And the neural network will operate with these statistics.
Copy that, thank you