genetic algorithm
TRANSCRIPT
Genetic Algorithm
S.H - Fall 2014
Introduction
GA is a method normally used to generate useful solutions to optimization problems.
Genetic algorithms belong to the larger class of evolutionary algorithms (EA), which generate solutions to optimization problems using techniques inspired by natural evolution.
InheritanceCrossover
DNA needs to be copied when cells divide, so that each new cell has a complete set of genetic instructions.
The cellular process of copying DNA is full of mechanisms that check and double check the construction of a new DNA molecule. But when changes, or DNA mistakes do occur, it is usually harmful or at best has no effect on the organism.
Mutation
Based on Darwin’s Theory
Population
Individual (Chromosome)
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
Randomly generated individuals
A previously saved population
A set of solutions provided by a human expert
A set of solutions provided by another heuristic algorithm
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
Bit strings (0101 ... 1100)
Real numbers (43.2 -33.1 ... 89.2)
Permutations of element (E11 E3 E7 ... E1 E15)
Lists of rules (R1 R2 R3 ... R22 R23)
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
Determines the fitness of each member of the population.
Performs the objective function on each population member.
The most simple fitness function can be the objective function.
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
Inheritance - Crossover
Mutation
Crossover
Mutation
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
A solution is found that satisfies minimum criteria
Fixed number of generations reached
Allocated budget (computation time/money) reached
The highest ranking solution's fitness is reaching
No improvement in solution quality
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
Decoding
011101001
010001001
10010010
10010001
Gen
etic
Alg
ori
thm
Creating Initial Population
Encoding
Applying Fitness Function
Genetic Operators
Termination
Decoding
Selection
Crossover
Mutation
Max x2 over {0,1,…,31}(Selection)
Roulette Wheel
A C
1/6 = 17%
3/6 = 50%
B
2/6 = 33%fitness(A) = 3
fitness(B) = 1
fitness(C) = 2
Max x2 over {0,1,…,31}(Crossover)
Max x2 over {0,1,…,31}(Mutation)