Machine learning in trading: theory, models, practice and algo-trading - page 1159
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 wrote some bullshit in the evening from nothing to do :)
One NS tries to identify cycles with the highest possible response, the second extrapolates them into the future (like cycles are easier to predict than the prices themselves)
I turned around, it's so funny... if at the intersection with zero or from the supposed haves to open
I wrote some bullshit in the evening from nothing to do :)
One NS tries to identify cycles with the highest possible response, the second extrapolates them into the future (like cycles are easier to predict than the prices themselves)
It's so funny... if you open it at the intersection with zero or from expected hauls...
What an inventor you are!
Maybe the results of this joint work of networks can be turned into a predictor and given to other networks/trees to be eaten?
I wrote some bullshit in the evening from nothing to do :)
One NS tries to identify cycles with the highest possible response, the second extrapolates them into the future (like cycles are easier to predict than the prices themselves)
it's so funny... if at the intersection with zero or from assumed haves to open
I wanted to suggest today in the discussion of Library: RL GMDH: an Expert Advisor based on the MO is good, but imho, first we need to visualize what the EA sees, i.e. we need an indicator, not the EA, but now I see that you have already thought of everything ))))
Well, if you have an indicator, now you should write an EA based on it and take your time to observe in the visualization mode
SZZ: I saw an indicator in kodobase not so long ago which looks exactly the same as yours, only it has no red lines on the right, is it not a frequency spectrum visualization?
What an inventor you are!
Maybe the results of this cooperative network work can be turned into a predictor and given to other networks/trees to eat?
redraws the future quite a bit, but overall this orange ball went to a contraction (big), and the small cycles predicted up and down well, after a while (the previous screenshot was a buy, for interest)
now again predicts a small red cycle upwards, i.e. growth of some small
i don't know, it's much more difficult to describe it with the help of a bot, but i can see it with my eyes
I put the test sticks to see if it turned around or not :) in general, the main signal - crossing the zero line
might be good for gambling
I put the test sticks to see if it turned around or not :) in general, the main signal - crossing the zero line
might be good for gambling
on the bottom you can validate (the greater the deviation from the middle the more accurate the result), less than some limit just do not believe,
and on the upper one (if the graph is correct, i.e. not involuntarily shifted to the future), then to enter by returns to the mean, i.e. a reversal downwards above 0 - sell, below 0 and upwards - buy.
PS/ (ok if not) but something tells me that one could (and probably someone did) come to these charts in a different way :-) Somehow, on a hunch, I see an autocorrelation of 4-5 bars and its estimation
on the lower one you can validate (the greater the deviation from the middle the more accurate the result), less than some limit simply do not believe,
and on the upper one (if the graph is correct, that is, it is not involuntarily shifted in the future), then to enter by returns to the average, that is, the turn down above 0 sell, below 0 and up - buy.
PS/ (ok if not) but something tells me that one could (and probably someone did) come to these charts in a different way :-) Just off the top of my head I see a 4-5 bars autocorrelation and its estimation
Autoregression :) not shifted. to the average does not work - takes it out periodically. It does not work at crossings of 0-line normal signals, but it redraws the last bar of course. If you can somehow determine with your eyes what is about to happen... then I don't think so on the robot) it's just fun.
R.S. The cycle was not predicted, the extrapolator was redrawn. Even on such cycles it doesn't work, let alone try to price
And today I looked - the big shit was predicted normally, and today it gave a buy signal and predicts more growth. Okay, I'm done spamming :) but modifications of the AR\Arim can be interesting - it's for those who have nothing better to do
Can anyone advise if there are any machine learning algorithms?
I need to reduce the number of trades.
I can't find anything good.
He says it's time to know the honor
I'll try to make my own extrapolator, a better one.