Subject: IJCNN2019 discussion - Spiking Neural Networks and [DNA, RNA, etc]
From: "Bill Howell. Hussar. Alberta. Canada" <>
Date: Thu, 25 Jul 2019 14:58:22 -0600
To: "Alice Parker. Spiking Neural Networks & DNA. USC. Los Angeles. USA"

I enjoyed our conversation at the restaurant in Budapest last ?Saturday?.   It's good to find someone also concerned about the absence of some concepts in the ANN community, and the ongoing problems of generating relevant [models, architectures] that can be of greater relevance to neuro-scientists as working tools.  I keep in mind your comment, which if I remember correctly, was about a research group finding mRNA migration to the ?synapses or dendrites?. 

My own interests are partially described in the list of mostly unpublished notes below my signature block.   The published WCCI2006 [overview, commentary] paper is probably the best starting point, but I prefer an earlier draft version of the paper even though it is unpolished.  I failed to get the time to do the mathematics and an example problem, so it fell far short of what I hoped to do.  The MindCode Manifesto" is a real mess of point form thoughts, in suspended animation until I can get back to it.

When I do get back to "MindCode", my first objective will be to build a small-scale working model of a spiking neural network that does NOT require [rate, timing (STDP)] encoding of the conventional sense, even those could be involved

It's funny that you were involved in Hava Siegelmann's DARPA program, as her call for proposals was a primary reason for a brief flurry of point form notes in the summer of 2016, after some outlining of ideas in "The MindCode Manifesto" draft notes of 2015 that was a false restart of work.  I was thinking more of providing some ideas to her as possible concepts for others to pursue given my own lack of time to do a real project.  However, time did not allow me to build a more compact set of brainstorming ideas that might have been of interest to Hava.

Hopefully I can achieve sufficient progress over the next year with my project in fundamental theoretic physics, so that I can leave it aside and turn to the "MindCode" (description and link to a point form, incomplete document below). 

I really hope that you can lead something in the area of [DNA, RNA]-based SNNs. 

Mr. Bill Howell
P.O. Box 299, Hussar, Alberta, T0J1S0
member - International Neural Network Society (INNS), IEEE Computational Intelligence Society (IEEE-CIS),
WCCI2020 Glasgow, Publicity Chair mass emails,
Retired: Science Research Manager (SE-REM-01) at Natural Resources Canada, CanmetMINING, Ottawa

MindCode -  thoughts to date 

Robert Hecht-Nielson's Confabulation theory

Robert Hecht-Nielsen "Confabulation Theory : The mechanism of thought"  Springer-Verlag, Berlin  245pp ISBN 978-3-540-49603-8  The book comes with a DVD video in two parts

You spoke of a "flatter Deep Learning", that better resembles the cortical six layer structures.  Not only does Hecht-Nielsen's Confabulation Theory provide that, he directly ties his concept for cognition in mammals to specific neurological assumptions.

I was so impressed with his machine-generated responses that I conducted a non-scientific survey of unwilling friends and victims of casual encounter in bars etc.  This oprovides several quotes that illustrate the biological basis of his thinking (of course, the book is really needed to get a solid idea of that) :

Ted Burger's hippocampal prosthesis

Ted's long-term project is fascinating to me, and accomplishes far more than I would have thought possible.  A real-time functioning system in rats' brains, with some degree of human testing, although I still wonder about the latter - is it real?  In any case, it is [hardware, real biological spiking]-based (OK - wires to the outside), and therefore in line with your own background.  A company was formed to pursue this, but Ted withdrew, possibly because they may have been trying to move too fast?
T.W. Berger, J.J. Granacki, V.Z. Marmarelis,A.R. Tanguay, S.A. Deadwyler, G.A. Gerhardt, "Implantable Biomimetic Electronics as Neural Prostheses for Lost Cognitive Function", Proceedings of IJCNN 2005, International Joint Conference on Neural Networks. Montreal, paper #1745 pp 3109-3114, 31 July – 4 August 2005.
I've lost track of recent papers, one of which I reviewed for Giacomo Borracci related to a special issue of Computational Intelligence magazine.  That wasn't accepted, although I was impressed and there were substantive advances over their previous work.
IEEE_CIM_Nov2016_19 p Song, Robinson, Wang, Marmarelis, Hampson, Deadwyler, Berger - Characterization of Complex Brain Functions with Sparse Nonlinear Dynamical Modeling.pdf

# end-email