Beruflich Dokumente
Kultur Dokumente
Questions
schema
a template
the new gene alphabet {0,1,*}
allows exploration of similarities among
chromosomes
represents all matching strings
Schemata
example:
the schema (* 1 1 1 1 0 0 1 0 0) matches two
strings:
(0 1 1 1 1 0 0 1 0 0) and
(1 1 1 1 1 0 0 1 0 0)
Q. which strings does this schema match?
(0 1 1 * 1 0 1 1 * *)
Schemata
matches 2r strings
r: no. of (*) in schema
a string of length l matched by 2l schemata
for strings of length l total 3l schemata
population size n: between 2l and n*2l schemata
may be represented
Schemata
Examples:
S1: (* * * * 0 * *) o(S1)=1 d(S1)=0
S2: (* 1 * * 0 * *) o(S2)=2 d(S2)=3
S3: (0 1 * 0 0 * 1) o(S2)=5 d(S2)=6
Schema Processing
let m(S,t) denote the expected no. of individuals
matched by schema S at time t, what is m(S,t+1)?
Assume:
genotype length: l
proportionate parent selection
one-point crossover
bitwise mutation
generational
Schema Processing
(Selection)
f (S )
m( S , t 1) m( S , t ).n.
j fj
m( S , t 1) m( S , t )
f (S ) j fj
f f
n
Schema Processing
(Crossover)
d (S )
p s 1 pc
l 1
where pc is the probability of crossing over
Schema Processing
(Crossover)
example:
Consider
A=0111000 matched by
S1=*1****0
S2=***10**
Assume crossover site=2
What happens to both schemata?
Schema Processing
(Mutation)
for a schema to survive mutation with pm
all specified positions must survive
a single position survives with probability
(1-pm)
schema survives with probability
(1-pm)o(S) which is approximately (1-pmo(S) )
ps 1 pm o( S )
Schema Processing
f (S , t ) d (S )
m( S , t 1) m( S , t ) 1 pc pm o( S )
f (t ) l 1
where f (t ) is the average fitness of the population at time t
The Schema Theorem