Optimization processes analogous to nature

from Wikipedia, the free encyclopedia
QS IT
 This article was due to content flaws on the quality assurance side of the computer science editorial added. This is done in order to bring the quality of the articles from the subject area of ​​computer science to an acceptable level. Help to eliminate the shortcomings in this article and take part in the discussion !  ( + )


Reason: It lacks clarity and structure - Schwatzwutz !?! 14:35, Aug 6, 2013 (CEST)

Optimization procedures analogous to nature are metaheuristics , the basic functionality of which is inspired by biological models.

For problems for which no algorithm is known that finds the global optimum in an acceptable time (or at all), heuristics are used in order to find a sufficiently good solution in a shorter time. Typical natural phenomena that are used as an idea are, for example, evolution , swarm intelligence , cooling and the immune system of vertebrates .

Evolutionary Algorithms

The idea for these algorithms comes from biological evolution , in the framework of which organisms adapt to environmental conditions; therefore algorithmic analogues of selection , mutation, and recombination are used to solve complex optimization problems.

The evolutionary algorithms include:

  • Genetic programming
  • Genetic Algorithms
  • Evolution strategies
  • Evolutionary programming

Swarm intelligent algorithms

Motivated by the behavior of swarms / colonies from biology (schools of birds or fish, colonies of bees or ants), a solution to the problem is sought. The capabilities of individual swarm elements are very limited. Only the interaction of many elements makes it possible to find a good solution in the problem area.

Examples include:

Simulated cooling

The basic idea is to simulate a cooling process, for example during annealing in materials science. After heating a metal, the slow cooling ensures that the atoms have sufficient time to arrange themselves and form stable crystals. This achieves a low-energy state, close to the optimum. This class of algorithms is also used in particular for complex optimization tasks that are difficult to categorize.

Variants of the basic idea can be found under