1
GENETIC ALGORITHMS
Ranga RodrigoMarch 5, 2014
2
EVOLUTIONARY COMPUTATION (EC)
3
INTRODUCTION TO EVOLUTIONARY COMPUTATION
• Evolution is this process of adaption with the aim of improving the survival capabilities through processes such as – natural selection, – survival of the fittest, – reproduction, – mutation, – competition and – symbiosis.
4
DNA, the molecular basis for inheritance. Each strand of DNA is a chain of nucleotides, matching each other in the center to form what look like rungs on a twisted ladder.
http://en.wikipedia.org/wiki/Genetics
5
A Punnett square depicting a cross between two pea plants heterozygous for purple (B) and white (b) blossoms. At its most fundamental level, inheritance in organisms occurs by passing discrete heritable units, called genes, from parents to progeny.[31] This property was first observed by Gregor Mendel, who studied the segregation of heritable traits in pea plants.[12][32] In his experiments studying the trait for flower color, Mendel observed that the flowers of each pea plant were either purple or white—but never an intermediate between the two colors. These different, discrete versions of the same gene are called alleles.
http://en.wikipedia.org/wiki/Genetics
6
EVOLUTIONARY COMPUTING (EC)• Evolutionary computing models the processes of
natural evolution.• It is a computer-based problem solving systems that
use computational models of evolutionary processes, such as natural selection, survival of the fittest and reproduction.
7
EVOLUTIONARY ALGORITHM PARADIGMS
•Search or optimization based on genetic evolution: natural selection. By creating a population of solutions and applying genetic operators such as crossover and mutation to evolve the solutions in order to find the best one.
Genetic algorithms
•Evolution of a large number of randomly created computer programs to create program that solves a high-level problem.
•Based on genetic algorithms where individuals are programs represented as trees.
Genetic programming
•Derived from the simulation of adaptive behavior in evolution (i.e., phenotypic evolution).•Mostly applied to real-valued representations.
Evolutionary programming•Modeling the strategic parameters that control variation in evolution, i.e., the evolution of evolution.•For real-valued representation.
Evolutionary strategies
•Similar to genetic algorithms, differing in the reproduction mechanism used. Used for optimization of multi-dimensional real-valued functions.
Differential evolution
•Models the evolution of culture of a population and how the culture influences the genetic and phenotypic evolution of individuals.
Cultural algorithms
•Initially “dumb” individuals evolve through cooperation, or in competition with one another, acquiring the necessary characteristics to survive.
Coevolution
8
GENETIC ALGORITHMS (GA)
9
INTRODUCTION TO GA• Genetic algorithms imitate natural optimization
process, natural selection in evolution.• Developed by John Holland at the University of
Michigan for machine learning in 1975.• Mostly for binary representations.
10
EVOLUTIONARY SEARCH PROCESS
Initiation: Selection of
initial population of chromosomes
Evaluation of the fitness of
each chromosome
Checking for stopping criteria
Selection of chromosomes
Applying genetic
operators
Creating a new
population
Presentation of the best
chromosome
11
Start
Initiation: Selection of initial population of chromosomes
Evaluation of the fitness of chromosomes in the population
Stopping criterion
Selection of chromosomes
Application of genetic operators
Creating a new population
Stop
Presentation of the “best” chromosome
No Yes
12
SELECTION (ROULETTE WHEEL)• The fittest individuals must
have the greatest chance of survival.
• Probability of being selected
http://www.edc.ncl.ac.uk/highlight/rhjanuary2007g02.php/
N
jj
ii
f
fp
1
13
GENETIC OPERATORS• Crossover: combination of genetic material
randomly selected from two or more parents.
• Mutation: process of randomly changing the values of genes in a chromosome.