Docstoc

lecture genetic-algorithms

Document Sample
lecture genetic-algorithms Powered By Docstoc
					Evolutionary Systems and
   Genetic Algorithms
 Creative Systems (DESC9176)
Genetic Algorithms

“Genetic Algorithms are good at taking
large, potentially huge search spaces and
navigating them, looking for optimal
combinations of things, solutions you might
not otherwise find in a lifetime.”

- Salvatore Mangano
Computer Design, May 1995
     From RB to Evolutionary
• In rule based:
  – set of rules
  – knowledge (start conditions)
• Set of solutions
• In evolutionary
  – generation
  – fitness
  – iteration
     Darwin’s Theory of Natural
             Selection
• Two parents
  – Generation
• Fitness
  – Life and death
• English peppered moth




                          http://biology.clc.uc.edu/Courses/bi
                          o106/nat-sel.htm
                        Genetic Algorithms


                                                              Search techniques


                   Calculus-based techniques                           Guided random search techniques                 Enumerative techniques


            Direct methods            Indirect methods        Evolutionary algorithms      Simulated annealing         Dynamic programming


Finonacci                    Newton               Evolutionary strategies    Genetic algorithms


                                                                  Parallel                        Sequential


                                                         Centralized    Distributed     Steady-state    Generational




                       web.umr.edu/~ercal/387/slides/GATutorial.ppt
          Evolution as search
•   Large space
•   Specify fitness
•   Means of generating solutions in space
•   Directed search of space
               Generation
• Creation of new designs

• Parents breed:               Crossover
  – Some from parent A some from parent B


• Anomalies:                   Mutation
  – New value for a variable
  – e.g. colour
               Fitness
• Means of measuring performance of
  individual
• Likelihood of reproduction
• Evaluation of behaviour variables
  described in terms of structure
                 Pseudocode
{
    initialize population;
    evaluate population;
    while TerminationCriteriaNotSatisfied
    {
        select parents for reproduction;
        perform recombination and mutation;
        evaluate population;
    }
}
Evolved Plants (Sims 1991)
    Evolving virtual creatures
• Karl Sims
  http://www.youtu
    be.com/watch?
    v=F0OHycypS
    G8
                       Examples
• Facial animation
   – http://www.site.uottawa.ca/~wslee/publication/ISUVR2006.pdf
• Travelling salesman problem
   – http://www.lalena.com/ai/tsp/
• Generating surfaces for Maya
   – http://projects.csail.mit.edu/emergentDesign/genr8/index.html
• Robot controllers
   – http://www.robertwrose.com/cg/learning/rose-evorobotics.pdf
• Generating textures
   – http://www.csdl.tamu.edu/~mong/finalproj.htm
Almost eight years ago ... people at
Microsoft wrote a program [that] uses
some genetic things for finding short code
sequences. Windows 2.0 and 3.2, NT, and
almost all Microsoft applications products
have shipped with pieces of code created
by that system.”

 - Nathan Myhrvold, Microsoft Advanced
Technology Group, Wired, September
1995
   Applying a GA to a problem
• Finding structure variables
  – Encoding genotype
• Finding behaviour variables
  – Developing fitness function in terms of
    genotype
• Define crossover and mutation
  – How many parents?
  – Random crossover?
  – Chance of mutation?
        Encoding a problem
• What are the structure variables?

• What do I need to be able to describe the
  design?

• Example: number string, bitstring
  Developing a fitness function
• What are the behaviour variables?

• How do I describe behaviour variables in
  terms of the genotype?

• Example: assignation of points
                Selection
• Better individuals get higher chance
• Chances proportional to fitness
     Problems suited to GAs
Domain                  Application Types
Control                 gas pipeline, pole balancing, missile evasion, pursuit

Design                  semiconductor layout, aircraft design, keyboard
                        configuration, communication networks
Scheduling              manufacturing, facility scheduling, resource allocation

Robotics                trajectory planning

Machine Learning        designing neural networks, improving classification
                        algorithms, classifier systems
Signal Processing       filter design

Game Playing            poker, checkers, prisoner’s dilemma

Combinatorial           set covering, travelling salesman, routing, bin packing,
                        graph colouring and partitioning
Optimization
             web.umr.edu/~ercal/387/slides/GATutorial.ppt

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:4
posted:6/11/2010
language:English
pages:19
Chandra Sekhar Chandra Sekhar http://
About My name is chandra sekhar, working as professor