Genetic Algorithm

Genetic Algorithm

PRESENTED BY- TAUSEEF AHAMD
M.TECH (COMPUTER SC. & ENGINEERING)
COMPUTER ENGINEERING DEPARTMENT
ZAKIR HUSSAIN COLLEGE OF ENGG. & TECH.
A.M.U, ALIGARH

Outlines

 A quick overview of GA
 Features of GA
 Various Methods of Population Selection
 Anatomy Of GA
 An example of GA

References….

 Adaptation in Neural and Artificial
Systems, John Holland, 1975.
 Genetic Algorithm in Search, Optimization and
Machine Learning, David E. Goldberg, 1989.
 C. Darwin. On the Origin of Species by Means of
Natural Selection; or, the Preservation of flavored
Races in the Struggle for Life. John
Murray, London, 1859.

A quick overview of GA

 Developed: USA in the 1970’s, by John Holland
 Holland’s original GA is now known as the simple genetic
algorithm (SGA)
 GA was inspired by process of biological evolution
 It is based on the Darwin’s theory of “survival of the
fittest” : the better individuals have better chance of
reproducing.

Features of GA

 Used to solve Hard problems
 Maintains a POPULATION of solutions
 Solutions are encoded as CHROMOSOMES
 REPRODUCTION creates a new population
members
 MUTATION and CROSSOVER occurs during
reproduction

Population Selection
 stochastically select from one generation to
create the basis of the next generation
 The requirement is that the fittest individuals
have a greater chance of survival than weaker
ones
 fitter individuals will tend to have a better
probability of survival and will go forward to
form the mating pool for the next generation

Various Methods of population
Selection
a) Roulette Wheel selection
b) Rank Selection
c) Tournament Selection
d) Elitism

There are many other methods, but we will
discuss briefly only these methods.

Roulette Wheel selection(Example)

 Fitness f(x) of individual No. 3 is the fittest and
No. 2 is the weakest
 Strongest individual a value of 38% and the
weakest 5%
 These percentage fitness values can then be used
to configure the roulette wheel
 Number of times the roulette wheel is spun is
equal to size of the population
 Each time the wheel stops this gives the fitter
individuals the greatest chance of being selected
for the next generation and subsequent mating
pool.
 Individual No. 3: 01000001012 will become more
prevalent in the general population because it is
fitter

Tournament Selection
 Provides Selective pressure by holding a
tournament competition among n individuals
 Best individual from tournament is one having
highest fitness, which is the winner of
tournament
 Tournament competitions and winner is then
inserted into mating pool

Tournament selection( Example)

Rank Selection
 previous selection will have problems when the
fatnesses differs very much
 For example, if the best chromosome fitness is
90% of all the roulette wheel then the other
chromosomes will have very few chances to be
selected
 first ranks the population and then every
chromosome receives fitness from this ranking
 The worst will have fitness 1, second worst 2 etc.
and the best will have fitness N(number of
chromosomes in population).

Elitism
 Copies the best chromosome to new
offspring before the mutation and crossover
 When creating a new population by crossover
or mutation the best chromosome might be
lost
 Forces GA to retain some numbers of best
individuals at each generation
 Has been found that Elitism improves the
performance significantly

An Example

 Simple problem: max x2 over {0,1,…,31}
 GA approach:
 Representation: binary code, e.g. 01101 13
 Population size: 4
 1-point xover, bitwise mutation
 Roulette wheel selection
 Random initialisation
 We show one generational cycle done by
hand

Genetic Algorithm

More Related Content

What's hot (20)

Similar to Genetic Algorithm (20)

Recently uploaded (20)

Genetic Algorithm