EP Present

You might also like

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 3

Evolutionary Programming

Initialize set of solutions randomly with in a given bound.


Evaluate its fitness function or objective function F(x)
Mutate using zero-mean normally distributed probability
distribution

x i( t 1 ) x i( t )
i F ( x ) i N i ( 0 ,1)

where
i and imust be tuned such that the square function
is positive.
After mutation operation , both parent and offspring
populations are combined together and the best N solutions
are probabilistically selected for the next generation.
Selection operation:
For each solution x(i) in the combined population, a set of
members are chosen at random from the combined population.
The fitness of x(i) is compared with that of each member of the
chosen set and a score equal to the number of solutions having
the worst fitness is counted.
All 2N population members are sorted according to the
descending order of their scores.
The best N solutions having the better score are selected.
If =2N deterministic selection else <2N procedure is
probabilistic selection
Since both parent and offspring population are combined before selecting
N solutions , the selection procedure preserves elitism.

You might also like