Scribd
Upload
25 views9 pages

Session 2

The document discusses genetic algorithms and their key concepts including populations, chromosomes, genes, alleles, selection, crossover, and mutation. It also describes how genetic algorithms work by starting with a random population and using the principles of natural selection to evolve better solutions over multiple generations until a stopping criteria is met.

Uploaded by

carap71147
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
25 views9 pages

Session 2

The document discusses genetic algorithms and their key concepts including populations, chromosomes, genes, alleles, selection, crossover, and mutation. It also describes how genetic algorithms work by starting with a random population and using the principles of natural selection to evolve better solutions over multiple generations until a stopping criteria is met.

Uploaded by

carap71147
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 9

SESSION 2

Genetic Algorithms
WHAT IS GENETIC ALGORITHM
 Based on the philosophy that evolution is an optimization process
 Based on natural selection and survival of the fittest concept

 Simulates evolution

 Robustness: Balance between efficacy and efficiency needed for survival in


different environments.
TERMINOLOGY IN GENETIC ALGORITHMS
 Population: The subset of all possible solutions to the problem
 Chromosome: One such solution

 Gene: A bit position in the chromosome.

 Allele: Value of the gene


TESTING PERFORMANCE
 Rastrigin function
)} where n = 2 and A =10
 Rosenbrock function

where, a=1 and b=100


WHAT IS GENETIC ALGORITHM
 Based on natural selection – similar to biological evolution.
 Uses 3 types of rules
 Selection
 Crossover
 Mutation
WORKING PRINCIPLE
 Start with a random initial population.
 Create new population by combining existing solutions. To create new
population the following steps are followed.
 Compute fitness score of each member of the current population.
 Select members based on the fitness scores.
 Produce children from parents by combining the vector entries of parents i.e.
crossover and/or by making random changes to the member itself i.e. mutation.
 Replace the current population by new population.
 Stop when one of the stopping criteria is satisfied.
CROSSOVER ISSUES
 Elitist strategy: The one with highest fitness is not touched i.e. Not used for
crossover. This is passed directly onto the next generation of solutions.
STOPPING CRITERIA
There may be various stopping criteria for the algorithm. For example
 When the algorithm reaches a certain value for generations.

 When a time limit is crossed.

 When a certain fitness limit is reached i.e. the best member in the current
population has higher fitness score than what has been specified.
 When there is no significant improvement in the objective function over a
time known as stall time limit.
ADVANTAGES OF GA
 Works well on complicated problems
 Easy to do parallel computing

You might also like