Building a trading system using digital low-pass filters - page 23

New comment
 
Mathemat:
I can't get into myself, my wife reinstalled the mra... I'll talk to you in the afternoon, OK, Konstantin?

Of course, as convenient as possible, nothing important, just a link there...interesting, but that's for me, in case you've already read it and discarded the idea.
Anyway, tell me what you think...and see you tomorrow, good luck.
 
Mathemat:
There seems to be some workaround to the problem of synthetic generation that does not involve the stationarity of the process. But this is only a genitive for now. We would have to think it over.

Somebody posted Bendat there, chapter 12 is interesting, just about non-stationary series. Though the author writes more about non-stationary series arising during rocket launch, but still...

 
In short, according to Bulashev it looks as if fat tails put an impenetrable ban on the possibility in principle of determining the fact of stationarity/non-stationarity itself - at least as applied to returns or their logarithms. Well, of course, Foreh is not a barrel of honey. <br / translate="no">

Yes this reasoning by Bulashev, looks like a description of quantum mechanics on the fingers. I gave a link to the inversion method. It can be used to draw reliable conclusions within the required precision.
 
bstone:
In short, according to Bulashev it looks as if fat tails put an impenetrable ban on the very possibility of determining stationarity/non-stationarity in principle - at least as applied to returns or their logarithms. Well, I see, Foreh is not a barrel of honey.

Yes this reasoning by Bulashev, looks like a description of quantum mechanics on the fingers. I gave a link to the inversion method. It can be used to draw reliable conclusions within the required accuracy.


Just a reminder, Bulashev wrote the book "statistics for traders". :) it is the same as quantum mechanics for dummies. :)

And the inversion method is no more suitable than any other.

 
NorthernWind:

As a reminder, Bulashev wrote the book "statistics for traders". :) it's the same as quantum mechanics for dummies. :)

And the inversion method is no more suitable than any other.


I know that Bulashev is the author of the noted book, but I don't see how that detracts from the possibilities of the inversion method. Explain why it is not suitable?
 
NorthernWind:
bstone:
NorthernWind:

As a reminder, Bulashev wrote the book "statistics for traders". :) it's the same as quantum mechanics for dummies. :)

And the method of inversions is no more suitable than any other.


I know Bulashev is the author of the noted book, but I don't see how that detracts from the possibilities of the inversion method. Explain why it is not suitable?


There is a subtlety here, two completely different points that I personally do not mix together in any way.

Firstly, the author of the book is well aware of how traders are far from exact sciences, and presents material in most accessible for traders form - for what he deserves credit for, despite the fact that there are some remarks.

Secondly, it was said that "the method of inversions is no more suitable than other methods. It was not said that "it is not suitable", just "it is no better". [further sprinkled in, for vanity].

 
Well, from all that has been brought up here about stationarity, the inversion method is the only one that can be applied to our problem (in my opinion). If you have any ideas about other methods, we need to share them with the public :)
 
In general, the inversion criterion (the one I know), as well as the series criterion, gives some insight into possible dependencies between data, and does not directly indicate stationarity/non-stationarity.
 
The series criterion applied to the m.o. values obtained on the partitions of the data series under analysis allows conclusions to be drawn about the stationarity of this series.
 
This is too strong a statement for such a weak power criterion.
1...161718192021222324252627282930...32
New comment