See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/262165276

Advanced genetic operators and techniques: an analysis of dominance &

diploidy, reordering operator in genetic search

Conference Paper · May 2008

CITATION READS

1 23

2 authors, including:

Anuradha Deshpande
The Maharaja Sayajirao University of Baroda
12 PUBLICATIONS   6 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Anuradha Deshpande on 05 May 2015.

The user has requested enhancement of the downloaded file.

 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008

ADVANCED GENETIC OPERATORS AND TECHNIQUES -

AN ANALYSIS OF DOMINANCE & DIPLOIDY,
REORDERING OPERATOR IN GENETIC SEARCH

ANURADHA SANJIV DESHPANDE & RAMESH BALABHAU KELKAR

Department of Electrical Engineering, Faculty of Technology & Engineering,
MAHARAJA SAYAJIRAO UNIVERSITY OF BARODA
Dandia Bazar, Vadodara 390001
INDIA

Absract
The paper explains use of low level advanced Genetic Algorithm operator DOMINANCE & DIPLOIDY for
getting new binary string structure in terms of heterozygous or homozygous expression of dominated &
recessive alleles. It also highlights role of schema representation (physical schema and expressed schema) on
population size, as an attempt to improve upon the robustness of simple GA’s. The paper has demonstrated
numerically pattern of selection of dominant and recessive alleles. It has suggested numerically the
probability of gene movement and gene disruption for various REORDERING OPERATORS LIKE
INVERSION, CYCLE CROSSOVER, and order crossover etc. Theory of reordering operator is tested and
verified upon string / schema representation. Thus the paper aims to analyze the advanced genetic operators
and techniques in GA search.

Keywords:
Advanced Genetic operators and Techniques, Dominance&Diploidy, Schema, Physical Schema, Expressed
Schema, Binary String, Population, Probability

1) Introduction:

Genetic Algorithm is theoretically and empirically
proven to provide robust search in complex spaces.
Random search algorithm has achieved increasing
popularity as researchers have recognized the
shortcomings of calculus based & enumerate
schemes. The genetic algorithm is an example of a
search procedure that uses random choice as a tool
to guide a highly exploitative search through a
coding of parameter space.
Genetic algorithm is different from more normal
optimization & search procedures:
1) GA’s work with coding of the parameter set, not
the parameter themselves. Parameter are coded &
encoded in the string structure,
2) GA’s search is from a population of points, not a
single point. Choice of fittest is from many choices.
avenues for further improvement of genetic
3) GA’s uses theory of probability rules in
Selecting from one set of solutions (a
Population) to the next & not deterministic
Rules.
4) GA’s use objective function information to
guide the search, not derivative or other
auxiliary information.
Above objective is achieved by machination of
these operators like reproduction, Crossover
& mutation. Every attempt is made to improve
upon the robustness of simple GA’s. Limitations
are due to current state of knowledge
concerning natural genetic mechanisms.
Despite these limitations, the abstraction,
analysis, & implementation of advanced
operators & techniques are the most fruitful
algorithms.

ISBN: 978-960-6766-58-9 27 ISSN: 1790-5109

 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008
Dominance, inversion, intra chromosomal
duplication, deletion, translocation, & segregation
are low level operators, while niche speciation,
multi objective optimization are high level
operators. literature survey Suggest work of:
1)Redundancy of genotypes as the way for some
advanced operators in evolutionary algorithms-
Simulation study by HALINA KWASNICKA ,
institute of Applied informatics, Wroclaw
University of Technology, Poland. It is aimed to
study effects of some advanced evolutionary
operators like neutral mutation & macro mutation.
2)At international conference on evolutionary
multi criteria optimization in year 2001 & 2003
Eckart Zitler has demonstrated multi objective
evolutionary optimization basic principles &
various algorithmic aspects like fitness assignment
& environmental selection .No development on
dominance & diploidy operator of advanced
genetic algorithm. The paper has demonstrated
dominance & diploidy operator & it’s Analysis.
2) Problem Formulation
2.1) Dominance & diploidy:
The primary mechanism for eliminating the conflict
of redundancy is through a genetic operator that
geneticists have called Dominance. At a locus, it
has been observed that one allele (the dominant
allele) takes precedence over (dominates) the other
alternative alleles (the recessive) at that locus. More
specifically, an allele is dominant if it is expressed
(it shows up in the phenotype) when paired with
some other allele. If we assume that all dominant
characters are 1’s & all recessive characters are 0’s.
At each locus the dominant gene is always
expressed & that the recessive gene is only
expressed when it shows up in the company of
another recessive. In the geneticist’s parlance we
say that the dominant gene is expressed when
heterozygous (mixed, 10-1) or homozygous (pure,
11-1) and the recessive allele is expressed only
when homozygous (00-0).
Diploidy provides a mechanism for remembering
allele and allele combination that were previously
useful & that dominance provides an opportunity to
shield those remembered alleles from harmful
selector, in a currently hostile environment.
ISBN: 978-960-6766-58-9 28
2.2) AN ANALYSIS OF DOMINANCE &
DIPLOIDY IN GA SEARCH:
Diploidy & Dominance were looked to as a
magic elixir to cure all GA ills; the focus is now
on their important role in shielding once
successful schemata from overzealous extinction.
It is likely that the combined action of Diploidy
and dominance prolongs the life of currently
weak, but once useful, alternatives & permit a
lower back ground mutation rate to maintain a
certain level of diversity. The effect of
dominance & diploidy is on the schemata growth
& decay.
The number of schemata H contained in the
next population written as m (H, T+1) is related
to the number in the current population m (H,
T) by the following equation:
f (H ) δ (H )
m( H , t + 1) >= m( H , t ) × [1 ∓ Pc ]
f av l ∓1
..........(1)
The number of schemata H contained in the
next population written as m (H, T+1) is related
to the number in the current population m (H,
T) due to dominance & diploidy operator is
given by the following equation:
f (H ) δ (H )
m( H , t + 1) >= m( H , t ) × [1 ∓ Pc
f av l ∓1
∓ o( H ) Pm ]..........(2)
Pc = crossover probability
Pm = mutation probability
f (H) = Schema average fitness
fav = population average fitness
δ (H) = defining length of the schema (distance
between outer most fixed position)
O (H) = Schema order (number of defined
position)
With the addition of dominance & diploidy, the
equation is still an accurate description of the growth
or decay of schemata if we recognize the effect of
dominance & allele expression on the schema
average fitness f (H). The difference becomes more
striking if we separate the physical schema H from
the expressed schema He. A real or physical schema
H may or may not be expressed, depending upon its
ISSN: 1790-5109
 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008
state of dominance & its current homologous
partner. This requires the following modification to
the schema growth equation:
f (H ) δ (H )
m( H , t + 1) >= m( H , t ) × [1 ∓ Pc
f av l ∓1
∓ o( H ) Pm ]..........(3)
Every thing remains same, except the average fitness
of the schema H, F (H) is replaced by the average
fitness of the expressed schema He, f (He). In the
case of a fully dominant schema H, the average fitness
of the physical schema always equals the expected
average fitness of the expressed schema He:
f (H) = f (He)
Above event is possible if parent strings are repeated as
offspring after dominance & diploidy application.
In the case of a dominated schema He, the hope is, of
course, that the average fitness of the expressed schema
is greater than or equal to the average fitness of the
physical schema:
f (He) >= f (H)
The situation is most likely to occur when the
dominance map is permitted to evolve. If this situation
does arise, then the currently deleterious, dominated
schema will not be selected out of the population as
rapidly as in the corresponding haploid situation.This
is how dominance & diploidy shield currently out of
favor schemata.
3) Problem Solution:
3.1) Implementation of Dominance & Diploidy
In the diploid form a genotype carries one or more
pairs of chromosomes, each containing information for
the same functions. As we have a pair of genes
describing each function.
Parent 1 - 1010011110100111
Parent 2 - 1101011011010110
Offspring 1 – 1111011111110111
Offspring 2 – 1000011010000110
Each position of a digit represents one allele. In the
above example, assume that digit 1 shows dominant &
digit 0 shows recessive gene.
ISBN: 978-960-6766-58-9
At any position or for each function, it has been
observed that one allele (the dominant allele)
takes precedence over (dominates) the other
alternative allele (the recessive) at that position.
More specifically an allele is dominant if it is
expressed when paired with some other allele.
At each position the dominant gene is always
expressed & that the recessive gene is only
expressed when it shows up in the company of
another recessive. In the geneticist’s parlance, it
is said that the dominant gene is expressed when
heterozygous (mixed, 1 0 – 1) or homozygous
(pure, 1 1 – 1) & the recessive allele is expressed
only when homozygous (0 0 – 0)
The redundant memory of diploidy permits
multiple solutions to be carried out along with
only one particular solution expressed. In this
way old lessons are not lost forever, and
dominated change permit the old lessons to be
remembered & tested occasionally.
Consider two strings forming physical schema:
String1–1010011110100111 (parent1)
String 2- 1101011011010110 (parent2)
After applying dominance & diploidy Operator
as described above, the resulting Offspring will
be the expressed schema:
Offspring 1 - 1111011111110111
Offspring 2 - 1010011110100111
Offspring 3 – 1000011010000110
3.2) Calculation of schema number:
For Parent string:
f (H ) δ (H )
m( H , t + 1) >= m( H , t ) × [1 ∓ Pc
f av l ∓1
∓ o( H ) Pm ]..........(4)
For parent strings 1&2, representative schemata
H1 = 1***011*1***011, schemata order = 2 &
defining length = 15.
The effect of reproduction, crossover &
mutation on the first schema H1 is observed as
under. During reproduction phase, the strings are
copied probabilistically according to their fitness
values.
29 ISSN: 1790-5109
 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008

TABLE NO 1 & 2
GA PROCESSING OF SCHEMA (PARENT)

S Initial X F(x)=x^2 Psel Exp Act

r population value ect e ual
N Fi/Σ cted cou
o f cou nt
nt
1 1010011110100111 42919 1842040961 0.38 0.76 1.0
2 1101011011010110 54997 3024670009 0.62 1.24 2.0

s 4866710970 1.0 2.0 2.0 CALCULATION OF PARENT &

u OFFSPRING SCHEMA RELATIONSHIP
m
a 2433355485 0.5 1.0 1.0
v m (H, t+1) >= m (H, t) f (He) / fav [1 – (Pc δ(H)/l-
g 1) – (O (H) Pm)]
m 3024670009 0.62 0.98 1.0 For Parent
a with Pc=0.009, Pm=0.001,δ(H)=15,O(H)=2, f(H)
x
=2433355485
m (H, t+1) =2
For Offspring
with Pc=0.009, Pm=0.001,δ(H)=15,O(H)=2,
Schema Processing
f(H)=2607745205
before Reproduction
m (H, t+1) = 1
String Schema
Representative Average fitness
f(H)
H1 1***011*1***011* 1,2 2433355485 3.3) Search for recessive allele
Both strings are representative of schema
Dominant heterozygous or homozygous is
TABLE NO 3 & 4 always expressed while recessive homozygous is
GA PROCESSING OF SCHEMA (OFFSPRINGS) expressed only when homozygous
rearrangement of schema growth equation
Sr No Initial X F(x)=x^2 permits calculation of recessive alleles, Pt, in
population value successive generation t+1.
1 1000011010000110 34437 1185906969 Assuming only two alternatives, Dominant form
2 1111011111110111 63479 4029583441 having constant expected fitness of fd & the
sum 5215490410 recessive fr of recessive expected fitness, the
avg 2607745205 proportion of recessive expected in next
max 4029583441 generation is calculated as follows:

P t +1 = P t × k × [( Pt ± r × (1 ∓ P t )) ÷
SCHEMA PROCESSING
AFTER REPRODUCTION ((1 ∓ r ) × P t × P t ± r ))]
String Expected Actual Schema
Representative count count Average
fitness where r=fd / fr, & K is crossover mutation rate
f(H)
H1 1***011*1***011 1,2 3.96 3.95 2607745205
3.4 ) RESULT OF DOMINANCE OPERATOR:

For offspring strings 1 & 2, representative schema H1 is Dominant fitness=1.0, recessive fitness=<1.0,
1***011*1***011*. for e.g. 0.991, 0.927, r=1.08, population=10,

ISBN: 978-960-6766-58-9 30 ISSN: 1790-5109

 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008
Recessive offspring’s = 9, Pt=9/10=>0.9,
K=1.008
Substituting in recessive Proportion formulae,
Pts obtained as 0.001,
A similar equation is derived for the haploid
case where deleterious alternative (recessive) is
always expressed when present in a haploid
structure:
P t +1 = [( P t × k ) ÷ ( P t ± r × (1 ∓ P t ))] =>
0.892
Dominance & Diploidy induce long term
memory. Because of this effect, under diploidy
& dominance we expect that mutation should
play an even smaller role in the operation of the
GA.
The proportion of recessive alleles in the next
generation Pt+1 is related to the proportion in
the current generation by the following equation:
P t +1 = [(1 ∓ ε ) × P t ± Pm × (1 ∓ P t ) ∓ Pm P t ]
We have the sum of three terms, the proportion
due to selection, the source of alleles from
mutation, & the loss of alleles from mutation.
The ε (t) factor is the proportion lost due to
selection and other operator losses. In above
case of Dominance & Diploidy only 1 out of 10
are selected i.e.1 /10 =>0.1
Substituting in above equation we get,
Pt+1=0.811
At steady state, Pt+1=Pt=Pss
Solving for Pss, the equation is
Pss = Pm / (2*Pm + ε) =>0.01
This equation suggests that the final steady state
proportion of alleles is directly proportional to
the mutation rate (with large ε and small Pm).
For a diploid structure under selection &
mutation, it may be shown that the proportion of
recessive alleles in the next generation is related
to the number in the current generation by the
following equation:
P t +1 = [(1 ∓ 2 × ε × P t ) × P t ± 2 × Pm × (1 ∓ 2 × P t )]
At steady state a relationship between the
required mutation rate & the steady proportion
of recessive alleles:
Pm = [(ε × Pss Pss ) ÷ (1 ∓ 2 × Pss )]
For small steady state proportion of recessive
alleles, Pss < 1, this equation suggests that the
mutation rate required to keep a certain
ISBN: 978-960-6766-58-9
proportion of recessive alleles available is
proportional to the square of the proportion. The
presence of the same proportion of allele in the
diploidy case does not mean that they will be
sampled as frequently. Although the same
proportion is in existence, they are sampled
much less frequently (as the square of the
proportion). This underlines the need for
occasional dominance changes so stored alleles
can be sampled in the current context.
4) THEORY OF REORDERING
OPERATOR IN GA SEARCH:
Frantz calculated probability distribution of
orderings of a given specificity and defining
length. He determined probability of movement
of a gene located at a particular locus.
Probability of a randomly ordered permutation
with ‘o’ alleles of interest has a defining length
exactly equal to ‘δ’ is given by:
δ −2 l
P{D = δ ) = (((l ∓ δ ± 1)( )) ÷ ( ))
δ −2 δ
Cumulative Distribution:
δ lδ ∓ oδ ± δ l
P ( D ≤ δ )((( )( )) ÷ ( ))
o l o
The probability of gene movement under
inversion for a gene at locus K on a string of
length l is as follow: P {gene moved} =
2 × [k (l ± 1) ∓ (k 2 ± 1)] ÷ l 2
For long string length: P {moved} =
2 × ( x ∓ x where x is the non dimensionalised
2)
locus x= k/l
This Asymptotic expression has a maximum of
P = 0.5 at x= 0.5. (Verified in Sample example).
This imbalance in movement probability can be
reduced by 1) By treating the chromosome as a
ring, with no end & no beginning, each locus is
equally likely to be moved under a single
inversion (Two point ring crossover) & 2) To
leave the operator alone & live with its locus
dependency.
Successful grouping of genes could mitigate
towards the string ends there by using this
inversion shadow to retard further disruption.
On the other hand unsettled grouping of genes
might seek out the string centre, there by
31 ISSN: 1790-5109
 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008
guaranteeing themselves a high probability of
movement.
The locus dependence is an exploitable side
effect of the operator. Holland recognized this
fact in his calculation of the probability of
schema disruption due to inversion.
δ (H )
P (disruption) = 2 × Pi × ( )[1 ∓ (
l ∓1
4.1) VERIFICATION OF
REORDERING OPERATOR
Inversion Operator
Parent 1 1010011110100111
Parent 2 1101011011010110
Schema Representation: 1***011*1***011*
Offspring 1 – 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1
Offspring 2 – 1 0 0 0 0 1 1 0 1 0 0 0 0 1 1 0
Schema Representation: 1***011*1***011*
Calculation of Probability of Gene Movement
P {Gene moved} =2*[k (l + 1) – (K^2 + 1)]/l^2,
k=x*l. x is the non dimensionalised locus
For k=5, 8, 11
P {Gene moved} =0.461, 0.555, 0.5
Calculation of P [Moved] =2*(x-x^2)
For x=0.3125, 0.5, 0.6875
P [moved] =0.4296, 0.5, 0.4296
Crossover Operator
Parent 1 1111111111111011
Parent 2 1001111111101010
SchemaRepresentation:1**11111111*101*
δ (H)=4, o (H)=15
Offspring 1 – 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1
Offspring 2 – 1 0 0 0 0 1 1 0 1 0 0 0 0 1 1 0
Schema Representation: 10*1111111101011
δ (H) =3, o (H)=3
Calculation of Probability of Gene Movement
P {Gene moved} =2*[k (l + 1) – (K^2 + 1)]/l^2,
k=x*l. x is the non dimensionalised locus
For k=1, 4, 8, 13, 14
P {Gene moved} =0.1172, 0.398, 0.398, 0.3203
For x=0.0625, 0.25, 0.5, 0.8125, 0.875
P [moved] = 0.1172, 0.375, 0.5, 0.305, 0.22
P { disruption }:
δ (H ) δ (H )
P (disruption) = 2 × Pi × ( )[1 ∓ (
l ∓1 l ∓1
For x=0.3125 & Pi =0.5,
P [Disruption] = 0.23
ISBN: 978-960-6766-58-9
Order Crossover
Parent - 1***011*1***011*
Offspring 1***011*111*0***
δ (H) =5,o (H)=15, x=0.5625 & k=x*l=9
P {Gene moved} = 2 × [k (l ± 1) ∓ (k 2 ± 1)] ÷ l 2
δ (H ) P {Gene Moved} = 0.555
)]
l ∓1 P [Moved] =2*(x-x^2)
ALL P [Moved] =0.492
δ (H ) δ (H )
P (disruption) = 2 × Pi × ( )[1 ∓ ( )]
l ∓1 l ∓1
For x=0.5625 & Pi=0.492
P [Disruption] = 0.23
Analysis of results
¾ In the analysis of above parents &
offspring generation, I.e., Reproduction &
Crossover or Mutation generation, the expected
count & Actual count has remained same. This
indicates that the schema gives the same number
of copies in the next generation. The test was
done on one type of schema H1
1**011*1***011*.
¾ The schema H1 1***011*1***011* was
selected in both generation so as to minimize
variations arising out due to different schema
types & retain the number of copies from one
generation to next generation.
¾ The average fitness of the schema H1
1***011*1***011* in case of reproduction I.e.,
parents generation is 2433355485 called as f
(H). While the average fitness of the expressed
schema f (He) I.e., incase of crossover & / or
mutation I.e. offspring’s generation is
2607745205.
¾ In case of a fully dominant schema H,
the average fitness of the physical schema
always equals the expected average fitness of the
expressed schema He. F (H) = F (He). This is
possible if out of a number of possible schemas
)]
any schema type represents the string structure
quite similar to parent & offspring. If the
number of copies in the next generation are
32 ISSN: 1790-5109
 9th WSEAS International Conference on EVOLUTIONARY COMPUTING (EC’08), Sofia, Bulgaria, May 2-4, 2008
same as parents I.e., the binary digit equivalent
[sum f (x) = X^2/n] remains same although the
string structure may be different.
¾ In the case of a dominated schema H,
the hope is, of course, that the average fitness of
the expressed schema He is greater than or equal
to the average fitness of the physical schema H.
f (He)>= f (H) .In the case cited above, the
average fitness of expressed schema f (He) is
2607745205 while average fitness of physical
schema is 2433355485, I.e.
f (He)>= f (H), 2607745205 >=2433355485.
¾ If this situation does arise, then the
currently deleterious dominated schema will not
be selected out of the population as rapidly as in
the corresponding haploid situation. The
unidentified string structure in parents &
offspring may not be retained in successive
generation due to smaller or reduced population
size. This is how dominance & diploidy shield
out of favor schemata.
5) CONCLUSION:
[1]Dominance & diploidy are advanced
genetic operators which can give new binary
string structures in terms of heterozygous or
homozygous expression of dominant &
recessive allele.
[2]The physical schema average fitness H is
lesser than the expressed schema average
fitness He, then dominance & diploidy has
shielded (not selected) dominant schema out
of population. Also population size may be
reduced due to shielding.
[3]The physical schema average fitness H is
equal to the expressed schema average
fitness than the dominant schema are always
expressed I.e. .retained in next generation.
The population size is also not reduced.
[4]The recessive number of population
combination can be evaluated with different
ISBN: 978-960-6766-58-9
View publication stats
formulas. It indicates the pattern of selection of
dominant alleles.
[5]Theory of reordering operator suggests the
probability of gene movement & gene disruption
for various operators.
[6]Probability of gene moved & gene disruption
for various operators like Inversion operator,
cycle crossover operator, & order crossover
operator was calculated for parent & offspring
strings.
[7]The gene moved probability was obtained
maximum of 0.5 at x=0.5.For x>0.5, P<0.5. The
maximum value of P=0.5 is not violated in any
case.
REFERENCES:
BOOKS:
[1]Genetic Algorithms in search Optimization &
Machine Learning –David E. Goldberg
PAPERS:
[1]TutorialsatInternationalconferenceonevolutio
narymulticriterionoptimization (EMO 2001 &
EMO 2003) By Eckart Zitzler.
[2]Redundancy of Genotype as the way for some
advanced operators in evolutionary Algorithm –
Simulation Study, by Halina Kwasnicka,
Professor of computer Science at
InstituteofAppliedinformatics,Wroclow
UniversityofTechnology,Poland.ScienceatInstitu
te of Applied informatics,Wroclow Universityof
Technology, Poland.
[3]ME Dessertation of Ms Falguni Bhavsar M
titled “Application of Advanced Genetic
Algorithm Operators for economic &
environmental load dispatch”May 2006.
33 ISSN: 1790-5109