FIR filter with minimum phase - page 13

 
avtomat:

Do you know the methods of identification? -I can see that you don't know them.

not known, too lazy to google - what's wrong with them?
 
Demi:
not known, too lazy to google - what's wrong with them?


There is nothing wrong with them.
 
Demi:

the scheme is strange and does not make sense to me - it is about forex. Forex is an interbank market. All participants are banks and central banks of countries. The central bank can act as both a liquidity provider and a liquidity consumer.

This scheme has a place if we are talking about a kitchen. There is a liquidity supplier and there are liquidity consumers

And the feedback I wrote about has a place at the interbank. It simply does not exist in the kitchen, so there is no point in drawing different variants of price "management" for the "kitchen" trade



No, this is a scheme for the exchange and ECN. For the kitchen it's practically the same. For the interbank, change it up a bit)

P.S. Of course, the same participant can supply liquidity at one moment and consume it at the other.

 
avtomat:

As a first step, let's do the following conversion :

I note that already this first step allows us, having set a PF type, to carry out its parametric identification.

Note that this scheme lacks the main thing - the input signal. Which for the market are two things: the flow of external information and transactions of the largest participants (interventions). Both signals act on the system similarly, but are applied at different points in the scheme, so I separate them. Without taking them into account, there is no point in identifying parameters, because the original premise, which is evident from the outline - the market as a closed system - is wrong. It is necessary to consider an open system with unknown input signal and unknown parameters of PF components and solve the problem of "blind" deconvolution. And who said it would be easy.


ZZY And not a word about FIR filters, by the way.

 
alsu:

What is lacking in this scheme is an input signal. Which for the market are two things: the flow of external information and the transactions of the largest participants (interventions). Both signals have a similar effect on the system, but are applied at different points in the scheme, so I separate them. Without taking them into account, there is no point in identifying parameters, because the original premise, which is evident from the outline - the market as a closed system - is wrong. It is necessary to consider an open system with unknown input signal and unknown parameters of PF components and solve the problem of "blind" deconvolution. And who says that it will be easy.


ZZY And not a word about FIR filters, by the way.



(If you want to know more about the "non-price information" box and pull the arrow to "participants"))
 
Avals:


add the "non-price information" box and drag the arrow to "participants"))

plus the "interventions" box - and directly to the quotations)))
 
And we also need a delay line, preferably a longer one, and a weighted one. It should be placed on the "quotes" and from the outputs a thick cobweb should be drawn to the participants. This will be, so to speak, the influence of technical analysis on decision making))))
 
yes, yes... and a bow on the side - for aesthetes ;)
 
alsu:
And we also need a delay line, preferably a longer one, and a weighted one. Let's put it on the "quotes" and draw a thick web from outputs to the participants. This will be, so to speak, the influence of technical analysis on decision-making))))

easily):

non-price information and quotes are fed into analysis buffers corresponding to different investment horizons and the size of the decision window to be analysed. The accumulation of necessary information in the buffer creates a delay in decision-making.

The "non-price information" includes insider information, fundamental data, regulatory rules, trading schedules, etc., etc. All information on the basis of which decisions to buy or sell are made apart from quotes.

This information is an input for many decision-making algorithms (analysis methods). The interventions are also based on some block of analysis and decision-making methods (and they are also through the common cup and the flow of orders through the market). So the "non-price information" and the change in the analysis and decision-making methods of the market participants over time are external.

 
Avals:

easily)):

Nda..... My kindergarten teacher would take my pencil and paper away from me on such occasions and send me for a walk outside....

Add eight more arrows, colour the squares in different colours and start formalising the mat model

Reason: