Chapter 8
Replacing and Transitioning

In the previous three chapters, your primary focus was on transforming the existing population of chromosomes into a better population using selection, crossover, and mutation.

In Chapter 5, ​Selecting the Best​, you isolated a collection of chromosomes to be used as parents. You learned that it’s important that your parents are both fit and genetically diverse—so you can avoid premature convergence. In Chapter 6, ​Generating New Solutions​, you learned how to take chosen parents and use them to create new solutions. In Chapter 7, ​Preventing Premature Convergence​, you learned about mutation and how you can use it to stop your algorithms from converging too early.

Each of these steps—selection, crossover, and mutation—accepted a collection of chromosomes and transformed the collection in some way to produce a new collection of chromosomes. At this point, you can divide your population into three categories: parents, children, and leftovers.

You have three groups of chromosomes from which to choose, and you need to combine them in such a way that your new population retains the strengths of old chromosomes while integrating the strengths of new chromosomes. Ideally, good selection, crossover, and mutation strategies would effectively capture the strongest traits of older generations and combine them to make stronger generations; however, the process of combining the products of all of these steps is crucial to the success of an evolution.

The process of combining the byproducts of selection, crossover, and mutation into a new population is known as reinsertion or replacement. Both of these terms mean essentially the same thing; however, you’ll learn about some of the minute differences later on in this chapter.

For the remainder of this chapter, you’ll explore several approaches to reinsertion and replacement by applying different approaches to a realistic problem: creating the best class schedule for your next college semester.