Genetic Search Algorithms

grandgoatΤεχνίτη Νοημοσύνη και Ρομποτική

23 Οκτ 2013 (πριν από 3 χρόνια και 10 μήνες)

149 εμφανίσεις

Genetic Search
Algorithms

Matt Herbster

Why Another Search?


Designed in the 1950s, heavily
implemented under John Holland (1970s)


Genetic search is intended to simulate
natural systems


Works best on continuous and discrete
combinatorial problems


It will only take eight minutes of your time

Definitions


Chromosome


Gene


Allele


Locus


Genotype


Phenotype


String


Feature, character


Feature value


String position


Structure


Parameter set

Characteristics


Reproduction


Crossover


Mutation


Rarely used

Mutation Operations

Generative

Swap Node

Swap Sequence

Destructive

Crossover Operations

Single point

Order based

Other Representations


Array, matrix


Tree


String of bits


Any other data structure

Implementation

1.
Start with an initial gene pool

2.
Generate successors (either randomly
or deterministically) to create the first
generation pool

3.
Each node is evaluated by a fitness
function and sorted accordingly

4.
Create new generations with the better
most likely to reproduce

What is the meaning of the
word better?


As determined by fitness function
(essentially a heuristic)


Nodes with desired genes are
predetermined


Can often approach local maxima rather
than the global optimal solution


Assisted by random
-
restart hill climbing

Variations


Genetic algorithms produce optimal
results for many problems, eventually …


Speciation


Two nodes will reproduce
only if closely related


Technique helps improve speed


Parallel populations


simulates physical
separation with possible migration


Applications


Traveling salesman problem


Drilling of printed circuit boards


Planning bus routes


Scheduling


Computer games


represent an
evolution of players’ strategies


Stock market trading


data fitting, trend
spotting, budgeting

Questions