Information Technology Reference
In-Depth Information
As you can see in Figure 5.7, these two individuals represent slightly dif-
ferent solutions because one of the mutated constants happened to be part of
sub-ET 1 and, as a result, a slightly different program is generated. Note that
the two mutations that occurred in C 2 have no effect whatsoever in the
a.
0123456789012345601234567890123456
**a//aa?a??461226*a+*?a?aaaa406961-[m]
C
= {0.139, -0.299, -1.024, -0.330,
0.510
, -1.864, 1.008, -0.712, -1.740, 1.552}
m1
m2
C
= {-0.986, -0.147, -1.113, -1.577, 0.210,
0.218
, 1.705,
-0.770
, 1.845, 1.954}
0123456789012345601234567890123456
**a//aa?a??461226*a+*?a?aaaa406961-[d]
C = {0.139, -0.299, -1.024, -0.330, , -1.864, 1.008, -0.712, -1.740, 1.552}
C = {-0.986, -0.147, -1.113, -1.577, 0.210,
-0.256
d1
d2
0.853
, 1.705,
-0.256
, 1.845, 1.954}
b.
Sub-ET 1
Sub-ET 2
a
a
0.210
a
a
a
a
-0.986
0.510
c.
Sub-ET 1
Sub-ET 2
a
a
0.210
a
a
a
a
-0.256
-0.986
Figure 5.7. Illustration of direct mutation of random numerical constants. a) The
mother and daughter chromosomes with their arrays of RNCs. b) The sub-ETs
encoded by the mother chromosome (before mutation). c) The sub-ETs encoded by
the daughter chromosome (after mutation). The nodes affected by this event of
mutation are highlighted. Note that, of the three point mutations, just one of them
(the substitution of 0.510 by -0.256) left its mark in the expression tree; the other
two are neutral as they are not being used to build the sub-ETs.
Search WWH ::




Custom Search