Back ... ... ... ... ... ... ... ... ... ... ... Send comments to review committee. ... ... ... ... ... ... ... ... ... ... ... Forward
Main thread: E-mailed Remarks from Alex Zenkin edited into six beads and referenced from: (:..:) (**) (*) (*) (*) (*) (*)
[AZ]
The same opinion as to the regular rectangular grids I had in May'2000.
But after our discussions (you, Alex Citkin, Ludmilla, Anton (by internet) and I) our opinion became to change. Now I believe that your persistence and scientific intuition and talent can lead to a new deep methodology of the CCG-technology usage. Any natural (real) system tends to serve its state (the homeostasis principle). Only periodical and quasi-periodical local processes are realized in such systems. We have suggested that the CCG-technique allows one to "catch" the regularities and to produce good, easy interpreted CCG-images of such the processes.
Any abnormalities in the processes lead to visual changes of initial CCG-images, i.e., CCG-technique allows to see some (any?) transient processes.
I think that there are a wide area of "non-number-theoretical" applications of the CCG-technique to any time series. We need only to formulate neatly such the problem and to have a corresponding DATA about the process under consideration.
Up to now we spoke about CCG-visualization in the form of 2D-matrixes (grids) with a fixed number of columns (modulus of CCG-image). But there is an important degree of freedom in that there is a variable number of "cells" in each row. The variation of the number of cells may realize a new technique of, say, EEG-experts in deciphering the EEG-DATA and to "catch" some non-periodical transient processes. This technique must be at least partially informal, and thus relying on visual acuity.
To see and to understand process changes the visualization of the state of the process is quite value semantic knowledge.
Moreover, I think now about the idea of CCG-visualization of parallel processes. For example, it is quite difficult to predict prices or Dow-Jones average by means of common statistical methods. But if we shall visualize some other (e.g., political, industrial, natural, etc.) processes (time series) we shall be able to see some new correlations of a " cause - effect" kind. It is not trivial, but I think it is possible.