Module 09: Genetic Programming - Evolving Programs, Not Parameters

What if the solution isn't a vector of numbers but a computer program? GP evolves tree-structured expressions, discovering symbolic regression models, control policies, and even novel algorithms. It's evolution applied to code itself.

Learning Objectives

1. Understand tree-based GP representation (functions and terminals)
2. Implement subtree crossover and mutation
3. Handle bloat with parsimony pressure and depth limits
4. Use gplearn for symbolic regression
5. Understand variants: linear GP, grammatical evolution, strongly-typed GP

Concept Explanation

Coming soon.

Code Examples

Coming soon.

Exercises

Coming soon.

Milestone Checklist

• Built a GP system for symbolic regression
• Discovered a formula from data using GP
• Understand and mitigated bloat
• Used gplearn on a regression problem