Genetic algorithm

Question 1

Describe the genetic algorithm quickly?

Answer 1

With the genetic algorithm, we encode some solutions as “individuals”. They have some fitness - how well they do (eg. cycle length in TSP).

The algorithm loop looks like this:

• Increase population by mating individuals (crossover, mutate).
• Let individuals compete (eg. sort list by fitness).
• Repeat until stopping conditions are met.

Question 2

Describe crossover roughly.

Answer 2

In the mating process, two parents are combined to make two new children.

Crossover is the method in which the material is exchanged. We set some point in the material from which to start (two point: and end) and exchange everything in this interval.

Question 3

How do single-point and two-point crossover differ?

Answer 3

In single points crossover, you set a point. The children swap everything from this point to the end.

In two points crossover, you define two points to make an interval. The children swap only these intervals.

Question 4

How does mutation work?

Answer 4

With some random chance, the genetic material of an individual/child is changed by some small amount.