Towards the Automatic Programming of NEPs - Foundations on Natural and Artificial Computation - page 302

Towards the Automatic Programming of NEPs
Emilio del Rosal, Marina de la Cruz, and Alfonso Ortega de la Puente
Departamento de Ingenierıa Informatica, Escuela Politecnica Superior,
Universidad Autonoma de Madrid
{ emilio.delrosal,marina.cruz,alfonso.ortega } @uam.es
Abstract. This paper shows the platform with which we implement
a general methodology to automatically design NEPs to solve specific
problems. We use CGE/AGE (a new genetic programming algorithm)
and jNEP (a Java NEP simulator), two applications we have previously
developed. This work is just a proof of viability. We are interested on
linking all the modules and generating the initial population. Building
this platform is relevant, because our methodology includes several non
trivial steps, such as designing a grammar, and implementing and using
a simulator. For this first proof we have choosen a well known problem
that other authors have solved by means of NEPs.
1 Motivation
The task of automatically writing programs can be seen as a search problem:
finding the best in a set of candidate programs automatically generated. Any
general search technique can be used to solve this problem.
Conventional personal computers are based on the well known von Neumann
architecture, that can be considered as an implementation of the Turing ma-
chine. A great effort is being devoted to the design of new abstract computing
devices, which can be seen as alternative architectures to design new families
of computers. Some of them, inspired in the way used by Nature to solve dif-
ficult tasks eciently, are called natural or unconventional computers .Someof
the natural phenomena inspiring these devices are: the role of membranes in the
behaviour of cells, the structure of genetic information, biological networks that
perform natural processes in parallel, and the way in which species evolve.
Any computer scientist has a clear idea about how to program conventional
(von Neumann) computers by means of different high level programming lan-
guages and their corresponding compilers, which translate programs into ma-
chine code. On the other hand, imagining how to program unconventional com-
puters is quite di cult.
This paper shows a new platform using evolutionary computation for the auto-
matic programming of a family of natural computers (Networks of Evolutionary
Processors or NEPs).
This work was partially supported by the R&D program of the Community of Madrid
(S2009/TIC-1650, project “e-Madrid”). The authors thank Dr. Manuel Alfonseca for
his help to prepare this document.