Under Construction @ Keele 2016 Volume 1 Issue 1 | Page 58

48
!

Figure 4. The parameter range window, this is used to select the range the values from the
breakpoint function will be scaled to

3.2.2. Genetic Operators
The user listens to the sound generated by each genotype, if none are suitable a
new random population is created. If a sound is discovered there are three options:
create a child population through mutation, create a child population through
crossover, save the genotype into temporary storage.
Mutation uses a single parent genotype to create a child population. The
parent is copied into a new population a number of times equal to the chosen
population size. Variation in the child genotypes is created by randomly selecting and
flipping bits in the bitstring. 1's are changed to 0's and 0's to 1's. The number of bits
which are flipped is determined by the mutation rate, this allows the user to select
the percentage of bits which will be flipped. A low mutation rate of around 5 – 20%
tends to result in minor variations from the parent, a rate of around 20 – 40% tends
to create child genotypes which will be recognisable as connected to the parent but
with some substantial differences. Above a rate of 40% the results become more
unpredictable, there may be audible traces of the parent in the child genotypes or
they may be entirely unrecognisable. The effect of a given mutation rate can vary
somewhat depending on a number of factors such as the selected processing
parameters, the range of values which can be assigned to each parameter and the
number of breakpoints in each function.
Crossover uses two parent genotypes to create a child population. Three