What You Learned

Genetic Algorithms in Elixir — by Sean Moriarity (55 / 101)

👈 Implementing Chromosome Repairment | TOC | Chapter 7 Preventing Premature Convergence 👉

In this chapter, you learned about the importance of crossover and how crossover helps genetic algorithms exploit in search. You learned about N-queens and how choosing an appropriate crossover strategy affects the outcome of your algorithms.

You also learned four different types of crossover strategies that can be applied to different genotypes. Single-point crossover is the most basic and is only useful with binary genotypes. Uniform crossover is a slightly more useful crossover strategy, but it doesn’t preserve order in permutations. Order-one crossover is a crossover strategy specifically for permutation genotypes. Whole arithmetic recombination is a basic strategy for real-value genotypes.

With these four strategies, you have the ability to solve a wide variety of problems with different genotypes. You also should have a basic understanding of what strategies will work with different genotypes.

Finally, you learned about other less common techniques for generating new solutions, including crossing over multiple parents and chromosome repairment.

In the next chapter, you’ll explore how selection and crossover alone can lead to problems in your algorithm, and you’ll see how mutation addresses these problems.