[galib] Quick question on mutation probability
a.m.l.liekens at tue.nl
Fri Jun 11 03:31:30 EDT 2004
The probability of mutating a bit string is bitwise, similar to
mutations in nature. pMut = 0.01 means that each bit will be flipped
with this probability. There is a strictly positive probability that a
string is not mutated with pMut=0.01, i.e. (1-pMut)^stringLength. There
is also a very small probability (pMut^stringLength) that all bits in
the bitstring are flipped. As a result, it is not true that "at leat" 2
bits out of 200 will be mutated, but "on average" 2 bits will be
mutated. Assuming a bitwise mutation probability thus gives a mutation
rate for bitstrings that is proportional with the bitstrings' length,
i.e. longer bitstrings will, on average, have more bits flipped with a
constant mutation rate.
>If pMut = 0.01 and my binary string has a length of 200 bits. That means
>that at least two bits are always mutating on every chromosome?
>There is no chromosome that exhibits no change?
>Isn't that bad?
>If I am correct, the pMut should be proportional to the string size.
More information about the galib