Skip to content
BENIBIX Learn

Evolutionary Algorithm

AI🧭 FoundationsAI Approaches

Also known as Genetic Algorithm, Evolutionary Computation, Evolutionary Search

A search that copies evolution. It keeps the best answers, mixes them and repeats.

Think of it like

It is like breeding the fastest dogs. Keep the best, let them have pups, and repeat.

Example

Engineers want a radio antenna that works well. The program tries thousands of odd shapes. It keeps the best and mixes them, round after round.

Why it matters

It works when you can score an answer but cannot work out the best one by hand. It needs no slope to follow, only a score.

Definition

An evolutionary algorithm finds a good answer by copying how living things evolve. It starts with a big group of random answers. A score says how good each one is. The best answers are kept, mixed together and changed a little at random. This makes a new group, and the steps repeat for many rounds until the answers stop getting better.
Score, keep the best, mix and change, then score again.

Common mistake

Thinking it is sure to find the best answer. It finds a good one, and a different run may find another.

Key takeaways

  1. 1
    A score for each. Says how good an answer is.
  2. 2
    Keep the best. Weak answers are dropped.
  3. 3
    Mix and change. New answers come from old ones.

Learn it properly