Predicting the future with Fourier transforms - page 26
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
Notice, I didn't suggest that;)
If you know precisely the function describing the trajectory of something, even a price, but do not know the parameters of this function, then the task of determining these parameters is precisely the scope of interpolation methods. Example: A missile warning system is built on the knowledge of the parabolic nature of ballistic objects' motion. It is possible to determine their motion parameters from two trajectory points.
Edik, don't go crazy, if you please. The Fourier series is just one of the series invented to solve a very important and necessary interpolation problem, i.e. the problem of determining the intermediate value of a function between interpolation nodes by its values in the interpolation nodes.
about interpolation ballistics has interesting posts https://www.mql5.com/ru/forum/118526/page5
+ you could try this - penultimate post https://www.mql5.com/ru/forum/138116/page21 + Alexei's formula
The stroke is roughly proportional to the root of the period. That is, if you denote the average hourly stroke as middle_H1, then
middle_period ~ middle_H1 * sqrt( period / H1 ).
Notice, I didn't suggest that;)
If you know precisely the function describing the trajectory of something, even a price, but do not know the parameters of this function, then the task of determining these parameters is precisely the scope of interpolation methods. Example: A missile warning system is built on the knowledge of the parabolic nature of ballistic objects' motion. It is possible to determine their motion parameters from two trajectory points.
A missile has a restriction on the rate of change in its trajectory, a missile cannot change direction abruptly within a certain time, and where are such restrictions on prices?
Let's go back to the old Fourier.
It is the same series designed for interpolation. Its only fundamental difference from all other interpolation series is that it cannot in principle help to judge the future, because its continuation in any direction leads to its own REPEAT. That is all. Any extrapolation based on the Fourier transform predicts a complete repetition of events that have already occurred. Groundhog Day.
Are you sure he can't? What if I said it could. If I say that even a human being can withstand accelerations of hundreds or thousands of g?
You know what I mean, a rocket with a mass of, say, 100 kg, flying at a speed of 200 km/hour, cannot turn around 180 degrees in an instant and fly on. Such a turn will take some time (which can be calculated reliably in this problem), but the market does not.
Let's go back to the old Fourier.
It is the same series designed for interpolation. Its only fundamental difference from all other interpolation series is that it cannot in principle help to judge the future, because its continuation in any direction leads to its own REPEAT. That is all. Any extrapolation based on the Fourier transform predicts a complete repetition of events that have already occurred. Groundhog Day.
And why should I judge the future? I would like to know at least the state of motion of the present. I agree that extrapolation from the points of the cumulative sinusoid is nonsense.
What matters to me personally is not the inflection points of these sinusoids, but their moments of phase change. the link to the picture above was given.
https://www.mql5.com/ru/forum/138142/page12