A Class of Computing Models Inspired by Spiking Neurons:

A Public Colloquium on Spiking Neural P Systems

by Francis George C. Cabarle

Algorithms and Complexity lab

Department of Computer Science

University of the Philippines Diliman

Abstract

*S**piking
neural P systems* (in short, *SNP**
systems*) are a class of parallel,
distributed, and nondeterministic
models of computation. SNP
systems are membrane models within the larger *membrane
computing* area (which is further
included into *natural computing*)
initiated by Gheorghe
Paun
in 1998. In
what follows, we mention briefly some of
the topics that will be touched in this
colloquium. Membrane computing is a
systematic assault on computations inspired by the living cell,
continuing the goal of natural computing of obtaining computing ideas
from (or even performing computations in) the natural world. In
the case of
SNP systems, the inspiration mainly comes from a specific cell called
a *neuron*.
Neurons
are placed on the nodes of a directed graph with *synapses*
as arcs between neurons. The spiking of a
neuron
using a
spiking
rule produces
a signal simply called the *spike*.
The
spiking neuron uses time not only as a background for information,
but as information support also:
e.g. a number *n *can
be encoded as the time interval between the times *t'
– t = n*,
*t' > t, *of
spike production of a neuron. This
idea of using time as support produced
a
menagerie of ways to introduce inputs or extract outputs, which
gave rise to
many SNP system variants. Other
SNP
system variants include other neuroscience ideas,
which are then
used
for computing purposes, e.g.
astrocytes or glial cells, neural stem cell division.
The
issues of computability (how powerful a computing model is) and
computational efficiency (how much time and space is required for
computations)
are
given:
SNP systems and their
many variants are known to be Turing universal (i.e.
they can simulate a universal Turing machine) and computationally
efficient (i.e. they can efficiently solve **NP**-complete
problems). The
colloquium aims to provide an introduction on these computing issues
(both
closed and open)
for the interested viewers and researchers in
fields related to computing.

Keywords: Natural Computing, Membrane Computing, Spiking Neural P systems, Models of computation, Turing universality

Date: 23 April 2015, Thursday

Time: 16:30 – 18:00

Venue: ERDT room, Department of Computer Science,

Velasquez St., UP Diliman, Quezon city

Faculty, students, and the general public are invited.

The abstract poster is attached as a PDF file in this page.

The slides are attached as a PDF file in this page.