Week 6

Genetic Algorithm
By

Dr. Fazeel Abid

Assistant Professor

The University of Lahore, Pakistan

Genetic Algorithm - Introduction
• Genetic Algorithm (GA) is search optimization technique based on the principles
of Genetics and Natural Selection developed by John Holland.
• Find optimal or near-optimal solutions to difficult problems
• Algorithm based on Darwin’s Principle “Survival of the fittest”
• Subset of a much larger branch of computation known as Evolutionary
Computation.
• Provide efficient & effective techniques for optimization in:
Structural Engineering Problems
Computer Science
Image processing and Pattern Recognition
Neural Networks
Genetic Algorithm - Introduction
• GAs, have a pool or a population of possible solutions to the given problem.
• These solutions undergo
Recombination
Mutation
 Producing new children (Offspring)
 The process is repeated over various generations.
• Each individual (or candidate solution) assigned a fitness value
• The fitter individuals are given a higher chance to mate and yield more “fitter”
individuals. (“Survival of the Fittest”)
• In this way “evolving” of better individuals or solutions over generations, till
reach a stopping criterion.
Genetic Algorithms “Good-Enough” & “Fast-Enough”
Large Set Problems Good Solution Fast
If GPS Navigation system, takes a few
Large set of problems, which are NP-Hard, minutes (or even a few hours) to compute
GAs efficient to provide usable near- the “optimal” path from the source to
optimal solutions in a short amount of destination. Delay in such applications is
time. not acceptable and therefore a “good-
Failure In Landscapes Problems enough” solution, which is delivered
“fast” is required.
Traditional methods starts from random
point and moving in the direction of the
gradient, reach the top of the hill. However,
complex problem like landscapes, made of
many peaks causes such methods to fail, as
inherent tendency of getting stuck at the
local optima.
Genetic Algorithms – Basic Terminologies

• Population is a subset of all the possible

(encoded) solutions to the given problem.
• Chromosomes is one such solution to the
given problem.
• Gene is one element of a chromosome.
• Alleles Each gene for the blood type can
have the values 0, 1, or 2.
• Phenotype is a specific feature or
characteristic of the child such as the color
of the eye, blood group inherited from
parents.
Genetic Algorithms – Basic Terminologies
• Genotype A set of alleles form a Gene-pair.
• Encoding  is a process of transforming from the phenotype to genotype .
• Decoding is a process of transforming a solution from the genotype to the
phenotype.
• Fitness Function takes the solution as input and produces the suitability of the
solution as the output.
• Genetic Operators alter the genetic composition includes:
 Selection
 Crossover
 Mutation
Genetic Algorithm - Structure
• Start with an initial population (Random or
heuristics) select parents from this
population (Solutions) for mating.
• Apply crossover and mutation operators
on the parents to generate new off-springs.
• These off-springs replace the existing
individuals (Solutions) in the population
and the process repeats.
• In this way genetic algorithms actually try
to mimic the human evolution to some
extent.
Genetic Algorithms – Population & Initialization

• Population is a subset of solutions in the current generation.

• Several factors considered when dealing with GA population are:
The diversity of population should be maintained otherwise it might lead to premature
convergence.
The population size should not be kept very large as cause a GA to slow down, while
smaller population might not be enough for a good mating pool.
• There are two primary methods to initialize a population in a GA:
Random Initialization 
Populate the initial population with completely random solutions.
Heuristic initialization 
Populate the initial population using a heuristic for the problem.
Genetic Algorithms - Fitness Function
• A function which takes a candidate solution to the problem as input and
produces as output how “fit” our how “good” the solution with respect to
problem.
• Calculation of fitness value is done repeatedly in a GA, so it should be
sufficiently fast.
• A slow computation of the fitness value can adversely affect GA and make it
exceptionally slow.
• In most cases the fitness function and the objective function are the same as the
objective is to either maximize or minimize the given objective function.
• However, for more complex problems with multiple objectives and constraints,
an Algorithm Designer might choose to have a different fitness function.
Genetic Algorithms - Parent Selection
• Process of selecting parents which mate and recombine to create off-springs for
the next generation.
• Parent selection is very crucial to the convergence rate of the GA as good
parents drive individuals to a better and fitter solutions.
• Extremely fit solution taking over the entire population leads to the solutions
being close to one another in the solution space causes a loss of diversity. 
• Taking up of the entire population by one extremely fit solution is known
as premature convergence and is an undesirable condition in a GA.
• Maintaining good diversity in the population is extremely crucial for the
success of a GA.
Genetic Algorithms - Crossover
• The crossover operator is analogous to reproduction and
biological crossover.
• More than one parent is selected and one or more off-springs
are produced using the genetic material of the parents.
• Crossover is usually applied in a GA with a high probability
 pc .
Crossover Operators
• Crossover operators are very generic and the GA Designer
might choose to implement a problem-specific crossover
operator.
One Point Crossover
 A random crossover point is selected and the tails of its two
parents are swapped to get new off-springs.
Multi Point Crossover
 A generalization of the one-point crossover wherein
alternating segments are swapped to get new off-springs.
Genetic Algorithms - Mutation
• Maintain and introduce diversity in the genetic population and is usually applied with a low
probability – pm.
• If the probability is very high, the GA gets reduced to a random search.
Mutation Operators
GA designer might find a combination of these approaches or a
problem-specific mutation operator more useful.
Bit Flip Mutation
 Select one or more random bits and flip them. This is
used for binary encoded GAs.
Swap Mutation
 Select two positions on the chromosome at random, and
interchange the values.
Scramble Mutation
 From the entire chromosome, a subset of genes is chosen
and their values are scrambled or shuffled randomly.
Inversion Mutation
 Select a subset of genes like in scramble mutation, but
instead of shuffling the subset, merely invert the entire
string in the subset.
Genetic Algorithms - Survivor Selection

• Determines which individuals to be kicked

out or kept in the next generation.
• Should ensure that the fitter individuals are
not kicked out of the population, while at the
same time diversity should be maintained.
• GAs employ Elitism means that the current
fittest member of the population is always
propagated to the next generation.
• The following strategies are widely used for
Survivor Selection:
Age Based Selection
Fitness Based Selection
Genetic Algorithms - Termination Condition
• The termination condition of a Genetic Algorithm is important in determining
when a GA run will end.
• Usually want a termination condition such that our solution is close to the
optimal, at the end of the run.
• Usually, keep one of the following termination conditions :
When there has been no improvement in the population for X
iterations.
When reach an absolute number of generations.
When the objective function value has reached a certain pre-defined
value.
Lets Recall GA Significant Elements
Fitness Score/Function/Value
• A Fitness Score is given to each individual
which shows the ability of an individual to
“compete”. The individual having optimal fitness
score (or near optimal) are sought.
Selection Operator: 
• The idea is to give preference to the individuals with
good fitness scores and allow them to pass there
genes to the successive generations.
Crossover Operator
• This represents mating between individuals. Two
individuals are selected using selection operator and
crossover sites are chosen randomly. Then the genes
at these crossover sites are exchanged thus creating
a completely new individual (offspring).
Lets Recall GA Significant Elements
• Mutation Operator: The key idea
1)Randomly initialize populations p
is to insert random genes in
offspring to maintain the diversity 2) Determine fitness of population
in population to avoid the 3) Until convergence repeat:
premature convergence. a) Select parents from population
b) Crossover and generate new
population
c) Perform mutation on new
population
d) Calculate fitness for new
• The whole algorithm can be population
summarized as: