seniorGenetic Algorithms
What is the building block hypothesis in Genetic Algorithms?
Updated May 16, 2026
Short answer
The building block hypothesis states that GAs work by combining short, high-quality gene patterns into better solutions.
Deep explanation
The building block hypothesis explains GA efficiency by suggesting that small, highly fit gene patterns (schemas) are discovered early and recombined over generations to form increasingly optimal solutions. These 'building blocks' survive selection and are propagated via crossover. The hypothesis also implies that disruption of these blocks (due to poor crossover or mutation) reduces GA performance.
Real-world example
In neural architecture search, good layer patterns are reused across architectures.
Common mistakes
- Assuming all crossover operations preserve useful building blocks automatically.
Follow-up questions
- What is schema in GA?
- How does mutation affect building blocks?