‹header›
‹date/time›
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
‹footer›
‹#›
• In this introductory talk, I would like to
address some important issues complimentary to those addressed in my workshop
paper.
• ETS stands for “evolving transformation
system”.
Freeman Dyson
has been professor of physics at the Institute for Advanced Studies in Princeton from 1953.
Note the emphasis on the need for radically
new insights.
I have been there myself.
Ironically, this comes from the pen of
the leaders of statistical pattern recognition.
If you want see how this applies to the
ETS model, bring it up during the question period.
This is what von Neumann and Turing were
engaged in during the last years of their life.
See, however, regular workshops SSPR,
ICGI, and many other.
Here is Robert Ledley’s example of the
generative grammar for the two classes of chromosomes: submedian (S) and
telocentric (T). A bottom-up parsing process is also illustrated by the tree.
This is mainly due to the lack of any adequate
concept of class (or class representation).
For example, non-linear functions are
not compositions of the linear ones.
The classical computational models do
not satisfy these requirements, which is why both Turing and Von Neumann have
proceeded to look for new models.
•Mention
the example of isotopes in chemistry
•Primitive
transformations are temporal units of action
We use non-bold t (small
Greek tau) to denote a generic transformation.
Add the corresponding class weight
schemes and you have all the class representations (at this level).
The initial sites of the primitive
correspond to the interface sites, while the terminal sites of the primitive
correspond to the terminal sites of the supertransform.
See our main paper, the link to which is
given in the workshop announcement (see the second link in it).
Scientists have not gained yet
any experience with such new “measurement” processes.
Once you realize this, it comes as a shocking thought:
How could conventional decision surfaces in
Rn have been accepted as meaningful/useful
class descriptions? They cannot
in principle convey anything of
scientific interest neither about the classes nor their elements?
Moving on to the ETS model, . . .
Only some of the features are listed.