Genetic Algorithms

1 / 17

# Genetic Algorithms - PowerPoint PPT Presentation

Genetic Algorithms. Vida Movahedi November 2006. Contents. What are Genetic Algorithms? From Biology … Evolution … To Genetic Algorithms Demo. What are Genetic Algorithms?. A method of solving Optimization Problems Exponentially large set of solutions Easy to compute cost or value

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.

## PowerPoint Slideshow about 'Genetic Algorithms' - ford

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

### Genetic Algorithms

Vida Movahedi

November 2006

Contents
• What are Genetic Algorithms?
• From Biology …
• Evolution
• … To Genetic Algorithms
• Demo
What are Genetic Algorithms?
• A method of solving Optimization Problems
• Exponentially large set of solutions
• Easy to compute cost or value
• Search algorithm (looking for the optimum)
• Very similar to random search?!
• Population- based
• We start with a set of possible solutions (initial population) and evolve it to get to the optimum
• Also called Evolutionary Algorithms
• Based on evolution in biology

Can we use the same idea to get an optimal solution?

From Biology …
• Charles Darwin (1859)
• Natural selection , “survival of the fittest”
• Improvement of species
Evolution

To implement optimization as evolution, We need

• Mapping features to genes, showing each individual with a chromosome
• An initial population
• Have a function to measure fitness

 same as what we want to optimize

• Implement and apply Reproduction
• Replace offspring in old generation
• Have an exit condition for looping over generations
Initial Population
• Representation of possible solutions as chromosomes
• Binary
• Real
• etc.
• Random initial population
• If not random  stuck in local optima
Recombination (crossover)
• Random crossover points
• Inheriting genes from one parent
Mutation
• Random Mutation Point
• Changing gene value to a random value
… to Genetic Algorithms

BEGIN /* genetic algorithm*/

Generate initial population ;Compute fitness of each individual ;

LOOP

Select individuals from old generations for mating ;

Create offspring by applying recombination and/or mutation to the selected individuals ;

Compute fitness of the new individuals ;

Kill old individuals ,insert offspring in new generation ;

IF Population has converged THEN exit loop;

END LOOP

END

Example
• http://www.rennard.org/alife/english/gavgb.html
References
• [1] Hue, Xavier (1997), “Genetic Algorithms for Optimisation: Background and Applications”, http://www.epcc.ed.ac.uk/overview/publications/training_material/tech_watch/97_tw/techwatch-ga/
• [2] Whitely, Darell (1995), “A Genetic Algorithm Tutorial”, http://samizdat.mines.edu/ga_tutorial/