Random Flow Theory and FOREX - page 55
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
Stationarity is a special case of non-stationarity. Markets are stationary at certain intervals, but you cannot get market reversals from these intervals, i.e. what you can get from non-stationarity. On this post I gave a link to Peters. I'm rooting for the correct methodological approach. Efficient markets are a dead end, i.e. you can grind for centuries with depo emptying. If initially markets are considered as dynamic non-linear systems and the theory available in this field is directed to recognition of some patterns with prediction ability - it is perspective in its basis, maybe you or I will fail, but not a deadlock, i.e. one day, someone will get it, but in GER never and nobody.
White is a special case of black. You can look at things that way too, if you want to.
wow, that's really something. I'll try to explain without allegories: with DFT you get a decomposition of the signal into its components and then you will NOT be able to tell that it is NOT composed of them because the components will give you the original signal as a sum.
Let's be more correct - you get an approximation of an infinite-dimensional space by a finite-dimensional one, satisfying a given accuracy on the interpolation segment. And now pay attention - THIS absolutely DOES NOT MEAN THAT YOUR MODEL IS 100 ADEQUATIVE. It means only that on the given segment the process can be described in this way with an acceptable error. No more... Du Yu Aderstand ? These are substantially different things. It's not sufficient for prediction - you don't know anything about the essential properties of the process.
Good luck.
Stationarity is a special case of non-stationarity. Markets are stationary at certain intervals, but you cannot get market reversals from these intervals, i.e. what you can get from non-stationarity. On this post I gave a link to Peters. I'm rooting for the correct methodological approach. Efficient markets are a dead end, i.e. you can grind for centuries with depo emptying. If from the beginning markets are considered as dynamic non-linear systems and the theory available in this field is directed to recognising some patterns, which have predictive ability - it is perspective in its basis, maybe you or I will fail, but not a deadlock, i.e. one day, someone will get it, but in GER never and nobody.
Simple pattern recognition is, as far as I understand, within a stationary process. And we have a non-stationary one, meaning that patterns can change. Here either the methodology for pattern recognition must work in a non-stationary process (I have no idea) or it must take non-stationarity into account. The second is clearer.
Or are you assuming that the patterns are in areas of stationarity? >> But there is no such thing.
Negroes are from two continents. You can tell them apart from the picture, they're different :)
And they only live on two continents ? ;)
Look at the article I linked above, they only left this (key) thesis from GER (IMHO of course).
It's not a GER - it's a postulate of technical analysis. But once again, that's not the point. either we like Gauss with a normal distribution or we don't. Among the postulates is another one, which is my favourite - "history repeats itself".
PS. There is no link.
And only live on two continents ? ;)
They live on one continent - where the insurance company is located.
Let's be more correct - you will get an approximation of an infinite dimensional space by a finite dimensional one, satisfying the given accuracy on the interpolation segment. And now pay attention - THIS absolutely DOES NOT MEAN THAT YOUR MODEL IS 100 ADEQUATIVE. It means only that on the given segment the process can be described in this way with an acceptable error. No more... Du Yu Aderstand ? It is not sufficient for prediction - you know nothing about the essential properties of the process.
Good luck.
Ay anderstand, didd yu note anderstud wat ay min? Sorrie for poor eksplanayshin, i min ze seim seing - yu arere ONLY interpolating...
It's not a GER - it's a postulate of technical analysis. But again, that's not the point. either we like Gauss with a normal distribution or we don't. Among the postulates is another one, which is my favourite - "history repeats itself".
PS. There is no link.
Here it is again: http://web.mit.edu/alo/www/Papers/JPM2004.pdf
Ai anderstand, didd yu note anderstud wat ai min? Sorrie for poor eksplanayshin, ai min ze seim sing - yu areli interpolating...
Quite understand what you mean - only the statement that a signal consists of a sum of components because that sum can be approximated - is incorrect ;)...
>> Good luck.
Quite understand what you mean - only the statement that the signal consists of a sum of components because this sum can be approximated - is incorrect ;)...
Good luck.
4=2+2. It may be 3+1, but in any case 2+2 is correct.
P.S. It's been a few years since I graduated from botany. But something has settled in....