What is estimation of distribution algorithm (EDA) vs GA?

Updated May 16, 2026

Short answer

EDA replaces crossover/mutation with probabilistic model learning.

Deep explanation

Unlike GA, EDAs build a probabilistic model of promising solutions and sample new individuals from it. This removes explicit crossover and mutation and instead focuses on learning variable dependencies.

Real-world example

Learning feature distributions in automated machine learning systems.

Common mistakes

  • Assuming EDAs still use crossover like GA.

Follow-up questions

  • Why use EDAs?
  • What is downside?

More Genetic Algorithms interview questions

View all →