Extra-Intracellular Transgenetic Algorithm

roomycankerblossomAI and Robotics

Oct 23, 2013 (3 years and 5 months ago)

62 views


Extra
-
Intracellular Transgenetic Algorithm



Marco César Goldbarg

DIMAp, UFRN

Campus Universitário


Lagoa Nova

Natal, Brazil, 59 072
-
〸0

Elizabeth Ferreira Gouvêa



Abstract

This paper introduces a new Computational
Evolutionary algorithm, EITA, whic
h uses extra
and intracellular flows. EITA was applied to the
Quadratic Assignment and to the Graph Coloring
Problems.

1

INTRODUCTION

Computational Transgenetics
(CT) is a metaheuristic that
brings the following ideas to the evolutionary algorithms
context:
To use exogenous and endogenous information
to interfere on the processes of formation and
modification of individuals of a given population; to use
the intracellular flow (Kargupta, 1997) to manipulate
individuals; to explore new processes of population
i
mprovement using transgenetic agents and competition
between agents and individuals; to guide the evolutionary
process allowing the occurrence of evolutionary jumps.
R
ecent researches show that genes and culture are
inherently linked. Thus, individuals evo
lve both by
anatomical and behavioral selection. The rules that cause
anatomical and behavioral elements to come together are
called
epigenetic
(Lynch, 1998).

To bring information to
the evolutionary process, CT uses transgenetic agents to
manipulate indiv
iduals.
A CT agent is composed by one
or more memes and an operative method that come from
epigenetic rules.
Memes
are the elements of cultural
concepts.
A meme, in this work, is any proposal to
construct a set or block of genes (building block). A
meme ca
n be obtained from a number of sources, such as
heuristics, etc. The transgenetic agents manipulate
individuals of a certain population that evolves and may
reinforce the whole process adding new information to it.
CT algorithms (Goldbarg & Gouvêa, 2000) a
re designed
to consider the intracellular and epigenetic contexts.

2

EXTRA
-
INTRACELLULAR
TRANSGENETIC ALGORIT
HM

The Extra
-
Intracellular Transgenetic Algorithm, EITA,
can be described as shown below. An EITA has an
underlying Genetic Algorithm (GA). The arrow
s
(

),check list signs (

) and asterisks (*) mark the
statements where intracellular manipulations, epigenetic
and the underlying GA steps occur. The meme base
required by EITA can be thought as a library containing
information about the problem and the in
stance.





Load a Meme Base





Design a set of agents according to the Meme Base


(*) Generate and evaluate an initial population




Repeat






Lad a subset f agents


(*) Select a set f parents

t generate ffspring

††††††



M a n ip u la t e s e n s it iv e p a re n t s a c c rd in g t 
e p ig e n e t ic ru le s


(* ) Crs s v e r


(* ) M u t a t in

††††

††


Liberate chrms mes that can be s et free
(end f agent lifetime)





(*) Evaluate ppulatin fitnes s






Until the stp criterin be satisfied



3

COMPUTATIONAL EXPERI
MENTS

In order to check the transgenetic potential, the approach
was applied to the Quadratic Assignment Problem (QAP)
and to the Graph Coloring Pro
blem (GCP).
Computational experiments showed that chromosome
manipulation by transgenetic agents is a powerful tool to
guide the search in the solution space.

References

M. C. Goldbarg, and E. Gouvêa (2000). Computational
Transgenetics.

X Congreso Latino
-
Iberoamericano de
Investigación de Operaciones y Sistemas
, Mexico.

H. Kargupta (1997). Gene Expression: The Missing Link
in Evolutionary Computation,
Genetic Algorithms in
Engineering and Computer Science
. John Wiley & Sons.

A. Lynch (1998). Units, Events
and Dynamics in
Memetic Evolution.
Journal of Memetics
-

Evolutionary
Models of Information Transmission,

2.