Demonstrating the cluster approach to market... - page 13

New comment
 
sab1uk >> :

>>mcd is a parody of bandpass filter, and mashka is a parody of low pass filter.

I can't see how a bandpass filter has anything to do with the market...

 

I showed you the software, you can do both low-pass and bandpass filters there.

The differential between the two LPFs is the same as the bandpass

We'll continue the master class tomorrow.

 
Mathemat >> :

I can't see how a bandpass filter has anything to do with the market...

>> So, the basis of spectral analysis!

 
sab1uk писал(а) >>

not markets but market cycles http://fx.qrz.ru/images/screen4.gif

there is a pitfall in this picture. Or rather in the method of obtaining this spectrum. The spectrum is actually different.

 
Prival писал(а) >>

there is a pitfall in this picture. Or rather in the method of obtaining this spectrum. The spectrum is actually different.

Well I said it's not a spectrum, it's a spectral density

the maximum entropy method is used to calculate matched filters

 
sab1uk писал(а) >>

Well, that's what I said, not spectrum, but spectral density

The maximum entropy method is used to calculate matched filters

No. It is not spectral density as it is defined. the MME contains in itself some model (this is an essence of this method). it is a spectrum of conformity of this model. the point is that if to an input of this algorithm give a white noise. it will find loops there too.

 

Yes, that's because:

Белый шум — стационарный шум, спектральные составляющие которого равномерно распределены по всему диапазону задействованных частот.

 
sab1uk >> :

Yes, that's because:

To put it in human terms, it goes like this:

White noise is a signal whose power of all spectral (frequency) components is the same.

Therefore, white noise cannot carry any useful information.

If, however, one amplifies the power of a frequency, one can isolate that frequency and,

thus obtaining useful information at that frequency.

 
I was also interested in Semyon Semenych's method at the time, but there were many nuances in its application. This method was a pain in the ass. I had to think about it and experimented for a long time. I applied my own method of analysing the interaction of a group of currencies and obtained a currency demand indicator. The result is impressive. It works with less resources on any timeframe. The group analysis rules, especially for small timeframes. Here is an example: Green - dollar, blue - pound, red - franc, white - yen, lilac - eur.
 


Picture

1...67891011121314151617181920...31
New comment