Discussion of article "Practical evaluation of the adaptive market following method" - page 3
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
In this case, it is historical data on a selected time period that is analysed, not a sample of data. The difference between sampling and historical data is that not all data are taken, but only randomly selected data from the total data set.
If you see specific errors in the article, please point them out. I am open to discussion.
Good afternoon, Dmitry!
Perhaps there are no errors in the calculations, I don't want to diminish your talents and abilities as an obviously radio engineer.
You just, I hope, do not understand what you are working with yet. What kind of function? What are you applying the Fourier transform to? For example, apply your favourite transform to the probability density function from the Fokker-Planck equation or the Schrödinger wave function. And we'll rejoice.
Once again I say - strengths in one area of physics do not yield results in another area.The course of radioelectronics, for example, at physics faculties, in my time took exactly 1 semester, after which admirers of Fourier transform continued to study on this speciality and reached great heights in radio engineering, these are respected people and I treat them with respect, but further we did not apply this transform ANYWHERE and never in calculations. This is where their knowledge is useless. Here we need abstract thinking.
I disagree with the formula for determining important frequencies
In your example, based on market data, we see this:
Although the dotted frequency is higher than the neighbouring frequencies, it is much lower than all those to the left (except the neighbouring one), i.e. those to the left are more frequent and therefore more important, and we should not filter them out.
Essentially we should take the frequency from 0 and up to the intersection with -40db. 0 is certainly not logical, and very close to 0, because we will have to take into account a very large number of bars. I.e. it is necessary to determine the lower frequency in some other way.
And if we assume that there may be a situation as in the article with noisy sinusoids:
I.e. if there really appears a dominant frequency/frequencies - then it is necessary to search for extrema comparing each point with all points, not only with neighbouring ones.
And in general, you have made an interesting tool, thank you.
I will, because I know.
You not only don't know, but also don't believe in what you write, because if you knew, you wouldn't write.
The author wrote a code and an article and got wise. He got a fee and came two hundred dollars closer to a million. But you and your stereotypes are not.