Piligrimus is a neural network indicator. - page 6
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
To observe how an indicator works online - throw the indicator in the tester window (when testing any Expert Advisor) and watch how it behaves
it doesn't work with your indicator...
Put it up here, I think a lot of people will be interested to compare it with mine.
>> I can't put it up here, I'm not the author.
>> better in person!
To observe how an indicator works online - throw the indicator in the tester window (when testing any EA) and watch how it behaves
I will try to use your indicator but it does not work like that.
I have a visualizer where everything is tested.
To Piligrimm, I have tested your indicator, the signals are not bad, it does not render too much, but it shows lots of false signals, how can I get around that?
Will this indicator not be an adjustment? Over time, the coefficients in the calculation formula may not be the same, so you will have to recalculate for the new data?
I have everything tested in visualizer...
To Piligrimm, I tested your indicator, the signals are not bad, it does not overpredraw, but there will be a lot of false signals in the flat, how can I get around that?
If i wanted to use this indicator as a fitter, coefficients in calculation formula may not be right with time, i should recalculate them according to new data?
I am working on the third module now. It should just compensate errors in signals (like in my article "The Principle of Superposition and Interference of Financial Instruments"). The error compensation will decrease false positives in the flat.
There won't be any adjustment, in the foreseeable future I suppose, the market won't move much beyond the range it was in last year, though nothing is eternal in nature. But in case of such a situation, we may introduce scaling coefficients, again similar to the article, but the coefficient should be used proportionally to the deviation from the current price range.
Piligrimm, more details about scalable coefficient, like in the article ? I'm dealing with multi-currency...can you tell me if you can make a chart of a group of currencies taking into account their scale to graph the movement of all groups, so far i managed to do something clearly using price increment (sorry for the off-topic)
Piligrimm, could you tell us more about scalable coefficient, like in the article ? I am dealing with multi-currency....can you tell us if it is possible to make a chart of a group of currencies taking into account their scale so that the chart reflects the movement of all groups, so far it was done clearly using price increment (sorry for the off-topic)
I have a suggestion: the curves that show statics (trend) can be separated out and drawn only when they are static, in my opinion this would be more informative.
And on which version of the indicator are you trying, the first or the last one? The problem may be that kosa's Expert Advisor for the signal of the first version, and when you connect the indicator of the second version you must use another buffer - 5, the pink signal on the pictures, it corresponds to the signal of the first version, although you can also use buffer 3, I think the result will be even better.
it does, the expert works with the first version...
that's right, the expert works with the first version...
>>: I'll send you an email, I can't post it here.
I have a suggestion: the curves that show statics (trend) can be separated and drawn only when they are static, in my opinion, it will be more informative.
Not only these signals are important, but also the correlation of others with them, for example, a breakdown of these signals by others. Besides, in order to analyze a signal as static or not, it must be compared with the situation on the previous bar, and such a comparison will delay the solution by one bar. I struggle to decrease as much as possible the lag of signals relative to the trend, so I will try to solve this problem in another way. I'm working on the third module now, it will compensate errors and considerably improve amplitude-phase characteristics of the filter for all signals.